# Claude Code and Codex Assist — full site content > Every public page of https://ccassist.dev in Markdown, concatenated. > Generated at build time from the published pages. 17 pages. # Your AI coding history, finally organized > Browse Claude Code and Codex sessions, view file diffs, track token usage & costs. Advanced analytics and workflow tools for AI-assisted development. Source: https://ccassist.dev/ Last updated: 2026-10-01 --- Claude · Codex · Copilot + more Privacy-First Cross-Platform git blame, but for AI conversations — find the exact session and prompt that changed any file, then browse, search, resume, and track usage across Claude Code, Codex, OpenCode, Grok & Copilot. [Install in VS Code](vscode:extension/agsoft.claude-history-viewer)[Install in Cursor](cursor:extension/agsoft.claude-history-viewer) Also in[Antigravity](antigravity-ide:extension/agsoft.claude-history-viewer)[Trae](trae:extension/agsoft.claude-history-viewer)[Trae CN](trae-cn:extension/agsoft.claude-history-viewer) or from the [VS Code Marketplace](https://marketplace.visualstudio.com/items?itemName=agsoft.claude-history-viewer) or [Open VSX](https://open-vsx.org/extension/agsoft/claude-history-viewer) Join 1,000+ developers from 20+ countries Claude Code and Codex Assist Sound on ## Everything you need, nothing you don't Features built for developers who use AI coding assistants daily. Session management, code tracking, and analytics in one extension. ### Five Assistant History Claude Code, Codex, OpenCode, Grok, and Copilot sessions with source badges ### Pin, Rename & Organize Pin important sessions, custom titles, smart filtering by project ### Fork & Resume Sessions Resume every supported source; fork Claude and Codex sessions ### Session Format Converter Convert sessions between Claude and Codex formats ### Session Archive Keep full-fidelity transcripts safe from CLI auto-cleanup ### Markdown Sync Export Markdown files you can commit to git and restore across devices ### GitHub-Style Diffs Beautiful, readable diffs with syntax highlighting across files ### Review Changes View See all file changes from a conversation in one organized view ### One-Click Undo & Reapply Instantly undo or reapply any file change with a single click ### Multi-Diff & Multi-Edit View changes across multiple files simultaneously ### File History Timeline git blame for AI conversations: auto-follows the active file, then drill into any session for the exact prompt and its diff ### Deep Search Search every session straight from the files, with selectable scope, match mode, tool output, and subagent options ### Cost & Token Tracking Token breakdown by type, per-model pricing, and cache-cost insights ### Usage Analytics Dashboard Activity heatmaps, daily/weekly trends, project-level breakdowns ### Quota Burn-Down Live quota cards, burn-down charts, weekly reset history, and status bar access ### Status Bar Navigation Quick navigation between file changes with keyboard shortcuts Cross-Platform (WebAssembly) Privacy-First — All Local ## A look inside Real screenshots from the sidebar. Everything you see runs locally on your own history. ### Find that one conversation in seconds Search every prompt and reply across thousands of sessions. Results come back grouped by session with the matching message, project, and cost, so you can jump straight to the moment it happened. * Pick who to search: everyone, only your prompts, or only the assistant * Match any words, an exact phrase, or a regular expression with one click * Optionally include tool output and subagent transcripts search ### git blame, but for AI conversations Open any file and the timeline lists every session that edited it, across Claude Code, Codex, OpenCode, Grok, and Copilot. Expand a session to read the exact prompt and open its diff. * Auto-follows the active editor, or right-click any file * Every session and prompt that touched the file, newest first * Jump from a prompt straight to the diff it produced timeline ### Review every edit, undo any of them Every turn shows the files it touched with +/- counts. Open a diff, undo a change, or reapply it. Review Changes gathers every file edit from a conversation in one place. * Per-turn file cards with +/- line counts * One-click diff, undo, and reapply for each edit * Review Changes: every file a conversation touched, in one view changes ### Every turn, every file it touched Open a session and read it like a transcript: rendered markdown, code, tasks, and a Turn changes card for each step. Resume it in the terminal, the desktop app, or the Codex extension, or export it as Markdown. * Per-turn file changes with +/- counts, diff, undo, and reapply * Context window graph, cache hit rate, cost, and turn count per session * Resume in Terminal, ChatGPT App, or the Codex VS Code extension conversation ![Conversation view with per-turn file changes and resume targets](/images/features/conversation-view.png) ### Know when you will run out Live 5-hour and weekly quota cards for Claude, Codex, Fable, and Grok, with burn-down charts that project when each window exhausts. An activity heatmap and a day-by-day cost table sit right below. * Exhausts in 2d 4h style projections from your real burn rate * Contribution-style activity graph with streaks * Tokens, cost, and session counts per day for the last 7 or 30 days quota ![Account tab with plan usage cards, quota burn-down, and activity heatmap](/images/features/account-quota.png) ### Let your agents read each other’s history A local MCP server exposes your session history to Claude Code, Codex, Grok, OpenCode, and Cursor. Say "continue my last Codex session" in Claude Code and it pulls a distilled handoff block. * search\_sessions, get\_context, get\_session, and list\_recent\_sessions * grep\_sessions, get\_session\_changes, and get\_file\_sessions on Pro * Loopback only, off by default, one-click setup for every CLI mcp ![MCP tab listing the ccassist tools available to CLI agents](/images/features/history-mcp.png) ### Every assistant, one list Claude Code, Codex, Grok, OpenCode, Copilot, and Cursor sessions grouped by project, with the cost of each folder next to its name. Filter by model or source, pin what matters, and resume from the row. * Per-model and per-source filters with session counts * Pin, rename, hide, or tag sessions to keep large histories tidy * Live dot on sessions that are still running sessions ![Sessions tab grouped by project with a model filter open](/images/features/chat-history.png) ### Your week, summarized A rolling 7-day view of sessions, tokens, cost, edits, and active days, each compared with the week before. Step back through earlier weeks to see how your usage is trending. * Day-by-day token bars with your heaviest day called out * Split by tool: Claude, Codex, Grok, Cursor * Top projects by cost and token volume digest ![Digest tab showing weekly totals, day-by-day tokens, and top projects](/images/features/weekly-digest.png) ### Quota at a glance, always Claude, Fable, Codex, and Grok percentages sit in the VS Code status bar so you never get surprised by a limit. Hover to see every window with its reset time. * 5-hour and weekly windows for each assistant * Reset countdowns without opening the sidebar * Click through to the full Account tab statusbar ![Status bar quota indicator with a hover popup listing each window](/images/features/usage-quota-statusbar.png) ## Simple, transparent pricing Get started for free. Upgrade when you're ready for more. ### Free $0forever All essential features at no cost * Browse all sessions (up to 7 days) * Indexed search (3-day limit) * File change tracking * Same-day diff viewing * Same-day undo/redo * Basic analytics (7-day data) * File timeline (3-day sessions) * 2 devices maximum [Get Started Free](/login/) ### Pro Monthly $4.99/month Unlimited history and up to 5 devices * Everything in Free, plus: * Unlimited diff viewing (all sessions) * Unlimited undo/redo (any session) * Session forking & pinning * Context window optimizer * Full file timeline (all sessions) * Advanced search (all results) * Full analytics & dashboard * Export sessions to Markdown * Up to 5 devices Start Pro Monthly Most Popular ### Pro Lifetime $49.99one-time Pays for itself in 10 months All pro features, forever — pay once * Everything in Free, plus: * Unlimited diff viewing (all sessions) * Unlimited undo/redo (any session) * Session forking & pinning * Context window optimizer * Full file timeline (all sessions) * Advanced search (all results) * Full analytics & dashboard * Export sessions to Markdown * Up to 5 devices * Lifetime updates included Get Lifetime Access ### Team Customper seat Pro seats for your whole engineering team * Everything in Pro, for every seat * One payment for all seats * Single invoice for your finance team * Volume pricing from 5 seats * Seats added or moved on request * Priority email support Contact us 7-Day Refund Policy: If Claude Code and Codex Assist doesn't work for your workflow, you can email me at [yash@ccassist.dev](mailto:yash@ccassist.dev) within 7 days for a refund. ## Three steps, zero config Install and start browsing your history in under a minute. 01 ### Install the extension One click from VS Code, Cursor, or Antigravity marketplace. Takes seconds. `$ ext install agsoft.claude-history-viewer` 02 ### Auto-detects your data Finds your ~/.claude/ and ~/.codex/ directories automatically. No configuration needed. `$ detecting ~/.claude/projects/ ... found 42 sessions` 03 ### Browse from the sidebar Sessions, diffs, search, analytics — everything accessible from VS Code's activity bar. `$ ready. open sidebar → CCAssist` ## Common questions ### Is this an official Claude or Codex product? No, this is an unofficial community extension created by independent developers. It is not affiliated with Anthropic or OpenAI. It enhances the Claude Code and Codex CLI experience. ### How does it access my data? Is it private? All processing is local by default. The extension reads from supported local assistant history folders such as ~/.claude/projects/ and ~/.codex/sessions/. No chat text, file content, or diffs are sent anywhere unless you explicitly opt in to community features. ### Which AI assistants are supported? Claude Code and Codex have full support for browsing, search, diffs, resume, branching from a message, and Claude-to-Codex or Codex-to-Claude format conversion. OpenCode, Grok, and GitHub Copilot support browsing, search, diffs, and resume; OpenCode requires VS Code 1.101+ with native SQLite support. ### Does it work on Apple Silicon Macs? Yes! The extension uses WebAssembly (sql.js) instead of native modules, so it works perfectly on M1, M2, M3, and M4 Macs — as well as Windows and Linux. ### What IDEs are supported? VS Code, Cursor, and Antigravity. Any editor that supports VS Code extensions should work. Install from the VS Code Marketplace or Open VSX Registry. ### What's included in Free vs Premium? Free includes 7-day session history, recent-only diffs/search/timeline, 3 export slots, 3 archive slots, and 2 devices. Pro is $4.99/mo or $49.99 lifetime and adds unlimited history, session fork, Claude-to-Codex and Codex-to-Claude format conversion, bulk export/delete, unlimited export/archive, and 5 devices. ### What is Session Archive? Session Archive copies full-fidelity session data into a separate archive folder so important transcripts can survive CLI auto-cleanup. Archived sessions remain browsable, and viewing or restoring archived sessions stays free; archiving new sessions beyond the free archive slots requires Pro. ### How does cost tracking work? The extension parses your chat history and calculates costs based on token usage and current model pricing. It provides accurate estimates with breakdowns by input, output, cache read, and cache creation tokens. Always verify against your actual billing. ### Can I get a refund? Yes. If the extension doesn't work for your workflow, email yash@ccassist.dev within 7 days of purchase for a refund. ### How do I report bugs or request features? Use the GitHub repository: github.com/yashagldit/Claude-Code-History-VSCode. Issues and pull requests are welcome. ## Ready to organize your AI workflow? Install in your favorite editor and start browsing your Claude and Codex history in seconds. [Install in VS Code](vscode:extension/agsoft.claude-history-viewer)[Install in Cursor](cursor:extension/agsoft.claude-history-viewer) Also in[Antigravity](antigravity-ide:extension/agsoft.claude-history-viewer)[Trae](trae:extension/agsoft.claude-history-viewer)[Trae CN](trae-cn:extension/agsoft.claude-history-viewer) or from the [VS Code Marketplace](https://marketplace.visualstudio.com/items?itemName=agsoft.claude-history-viewer) or [Open VSX](https://open-vsx.org/extension/agsoft/claude-history-viewer) --- # Claude Code and Codex CLI guides > How to find old Claude Code and Codex CLI sessions, what they cost you, and which VS Code extensions actually help. Written from using both tools daily. Source: https://ccassist.dev/guides/ Last updated: 2026-10-01 --- I built Claude Code and Codex Assist because I kept losing track of what the CLI had done the day before. These guides are the notes from that: where the session files are, how to get an old session back, how much the tokens cost, and which extensions I'd actually install. Each one says when it was last checked, because the CLIs change often. * Guide ## [Claude Code weekly limit and 5-hour limit, explained](/blog/claude-code-weekly-limit/) How the Claude Code weekly limit and 5-hour session limit work on Pro and Max, when each resets, how to check what's left, and what to do at the limit. * How-to ## [How to undo Claude Code changes (rewind, git, recovery)](/blog/undo-claude-code-changes/) Undo Claude Code changes with /rewind checkpoints, git and VS Code Local History, and recover files a Bash command deleted. Covers Codex CLI too. * Guide ## [Where Claude Code stores conversations (and why they vanish)](/blog/where-claude-code-stores-conversations/) Claude Code saves each session as a JSONL file in ~/.claude/projects and deletes it after 30 days. Exact paths per OS, the file format, and how to keep them. * Roundup ## [Best Claude Code VS Code extensions in 2026](/best-claude-code-vscode-extensions-2026/) The Claude Code VS Code extensions worth installing in 2026, grouped by job, with install counts and update dates checked on 1 October 2026. * Comparison ## [CCAssist vs ccusage: which Claude Code usage tool?](/compare/ccassist-vs-ccusage/) ccusage is a free CLI for Claude Code and Codex usage reports; CCAssist puts usage beside chat history and diffs in VS Code. Tested side by side on one week. * Feature guide ## [Claude Code usage tracker: tokens, cost per day, quota](/claude-code-usage-tracker/) How Claude Code usage is measured: the four token fields in each session file, how cost estimates use API prices, what /usage shows, and a daily cost script. * Feature guide ## [Codex CLI history: find, resume and read old sessions](/codex-cli-history-viewer/) Where Codex CLI keeps session history, how codex resume and codex fork find it, jq one-liners for rollout files, and viewing it all in VS Code. * How-to ## [How to view Claude Code chat history](/how-to/view-claude-code-chat-history/) View Claude Code chat history: claude --continue, --resume and the /resume picker, jq one-liners for the JSONL files, and searching every project in VS Code. ## Short version: where the history files are Claude Code keeps one JSONL file per session in `~/.claude/projects//`. Codex CLI keeps them in `~/.codex/sessions/YYYY/MM/DD/`. Nothing is encrypted or uploaded, they are just text files, so you can grep them yourself if you don't want an extension. The how-to guide covers both layouts and the`--resume` pickers if you only need to get back into yesterday's session. ## Other pages New posts appear first on the [blog](/blog/), which also has an [RSS feed](/blog/feed.xml). Release notes are in the [changelog](/changelog/). Features, pricing and install links for VS Code and Cursor are on the [home page](/). --- # Changelog > Recent Claude Code and Codex Assist releases, including File Recovery, the ccassist MCP server, session tags, Cursor support, Session Archive, and usage tracking updates. Source: https://ccassist.dev/changelog/ Last updated: 2026-09-27 --- These are the recent Claude Code and Codex Assist releases that changed assistant support, history browsing, file recovery, search, archiving, the ccassist MCP server, and usage analytics. The latest release listed here is 0.7.5 from September 27, 2026. ## What changed in 0.7.5? Version 0.7.5 was released on 2026-09-27. ### Added * File Recovery - A new Recovery tab finds saved copies of project files from Claude, Codex, and Grok sessions. Compare a copy with the current file or restore it, and filter by project, file status, or time. Today's copies are available on Free; Pro unlocks older copies. * Conversation Checkpoints - Claude and Grok conversations now show saved file states beside the prompts they belong to, so you can compare those files with their current versions. ### Improved * Clearer Conversation History - Messages exchanged between Claude sessions now show who sent them, while queued messages appear in the right place. Temporary scratchpad files are tucked away so they do not inflate file-change totals. * Easier History Browsing - Session rows have clearer highlights, long titles are easier to read, and compact rows reveal more details on hover. * Better Skills and Memories Filters - Narrow skills by source and memories by project directly from their tabs. * More Accurate Cost Estimates - Updated model prices now refresh older session costs, keeping history and usage totals current. ## What changed in 0.7.4? Version 0.7.4 was released on 2026-09-22. ### Improved * More Accurate Cursor Conversations - Cursor sessions now reconstruct user prompts, dynamic and MCP tools, tool results, errors, background-task notifications, timestamps, and file changes more reliably and in the correct order. * Smoother Live Conversation Following - Conversation views now stay pinned to the newest live output while messages, diffs, images, and visualizations expand, without pulling you back down after you intentionally scroll away. * Better Conversation Navigation - Mouse back buttons now return from a conversation to history, while links to agents and activity reliably scroll to the requested content. ### Fixed * Codex Subagent Sessions - Spawned Codex subagents no longer lose their own turns around compaction, and failed turns with no model request no longer inflate usage totals. * Tool Results in Activity Summaries - Tool results now remain attached to the correct activity even when a file-change card appears before the result arrives. ## What changed in 0.7.3? Version 0.7.3 was released on 2026-09-14. ### Added * Complete Support for Long Codex Sessions - Long Codex threads that continue across multiple pages now appear as one complete session everywhere, including history, live updates, search, resume, fork, conversion, archive, export, and ccassist MCP. * Inline Codex Visualizations - Interactive visualizations created by Codex now render directly inside the conversation view, with a clear fallback when the original visualization is no longer available. ### Improved * More Accurate Codex Usage and Costs - Usage analytics now avoid counting repeated token reports, include cache writes, and account for Fast-tier requests so totals more closely match actual Codex usage. ## What changed in 0.7.2? Version 0.7.2 was released on 2026-09-13. ### Added * Resume OpenCode Sessions in OpenChamber - OpenCode sessions can now open directly in the OpenChamber VS Code extension, with a new default resume preference and safe project-window switching when needed. ### Improved * Smarter ccassist Context Transfer - Choose how many recent turns to include when continuing a session, with better preservation of long messages and clear notices when content is shortened. * Faster Analytics and Share Cards - Usage views now read only the session details needed for the selected period and reuse project data while rendering. * More Reliable Live Session Updates - Newly created and updated sessions stay visible during background scans, and live status continues updating even when the search index is unavailable. * More Accurate Codex and Grok Sessions - Improved child-agent discovery, tool display, message counts, token totals, task updates, and session duration reporting. * Clearer Device Management - Linked devices now show more descriptive editor names, making similar installations easier to identify. ### Fixed * Worktree Project Grouping - Project headers, session rows, and “show more” results now stay together for removed or unusually located worktrees. * Quota Burn-Down Forecasts - Idle time is now included in burn-rate calculations, preventing overly pessimistic quota exhaustion estimates. ## What changed in 0.7.1? Version 0.7.1 was released on 2026-09-05. ### Added * Weekly Digest Tab - A new side panel tab summarizes your last 7 days of AI coding: sessions, tokens, cost, active days, a day-by-day chart, and your top tools and projects, compared against the week before. Step back through earlier calendar weeks, and get a once-a-week reminder when a fresh digest is ready. * Share Card - Turn today's, this week's, this month's, or a single project's usage into a shareable image you can copy or save. Everything is rendered locally from your own history; nothing is uploaded. * Branch and Worktree Filter - Narrow the history list to sessions from a specific git branch or worktree, alongside the existing source and model filters. * Activity Summary per Turn - Conversations now show what each turn did at a glance: commands run, files read, searches, and tools used, with collapsible detail rows and expand/collapse-all controls. * Reasoning Effort in Analytics - Session analytics and the day-by-day dashboard breakdown now split usage by model and reasoning effort, including thinking tokens, for Claude and Codex sessions. * Inline Local Images - Images referenced by local path in transcripts now render inline in the conversation view. ### Improved * Faster Startup and Sidebar - The extension starts faster with a compiled code cache, the sidebar re-renders less and reuses more of its UI, redundant file watchers were removed, and live polling pauses while the window is unfocused. * Session Notices When Resuming - Resuming a session that is already running in a terminal now tells you so and offers to fork instead. * Clearer System Events - Conversation system events now offer a "show result" link, and task notifications are handled more consistently. ### Fixed * Streamed Message Token Totals - Token usage for streamed assistant messages is no longer counted twice in conversation and analytics views. * Stale File Changes in Live Sessions - Live session refreshes now keep previously loaded file changes instead of briefly dropping them. * Effort Grouping - Reasoning effort values that differ only by case no longer split into separate rows in usage tables. * Experiment Toggles After Updates - Turning an experiment on or off right after an in-place extension update now prompts for a window reload instead of failing silently. ## What changed in 0.7.0? Version 0.7.0 was released on 2026-08-31. ### Added * Revamped Side Panel - The side panel has been redesigned with tab-based navigation, making it faster and cleaner to switch between history, account insights, settings, hooks, MCP, skills, memories, plans, and experiments. * Account Tab - Check plan usage, quota burn-down, recent activity, daily usage, account status, and device details directly from the side panel. * Real-Time Hooks Tab - Set up and manage live session status for Claude, Codex, and Grok, so the history list can show when an agent is working, waiting for permission, or finished. * Experiments Tab - Try upcoming features and vote on whether they should stay, including a customizable quota health bar, hidden session tags, and quicker Account access from the status bar. * Activity in Other Projects - When viewing one project, a banner now alerts you to active or unread sessions elsewhere and takes you back to the full list. * Per-Model Usage Breakdown - Expand each day in the analytics dashboard to see usage and cost totals for individual models. ### Improved * More Complete File Change History - File changes made through shell commands, PowerShell, loops, glob patterns, and subagents are detected more reliably across supported assistants. * Faster File Change Views - Large sessions and large file changes now open with less processing and memory use, while unchanged sessions reuse previously loaded results. * Cleaner History Controls - Project, source, model, sorting, grouping, and display controls are more compact and easier to navigate, with project actions available even when viewing a single project. * Cleaner Conversation Copying - Copied messages keep useful formatting without carrying VS Code theme colors or backgrounds into other apps. ### Fixed * File Change Accuracy - Reduced duplicate, incorrectly attributed, and missing file changes, including Windows paths, failed shell commands, and overlapping agent activity. * Codex Quota Cards - Codex usage cards now adapt to the windows available on the current plan instead of showing empty statistics. ## What changed in 0.6.9? Version 0.6.9 was released on 2026-08-21. ### Added * Offline Premium Activation - Premium can now be activated with a code from ccassist.dev, so firewalled or restricted editor environments can unlock Premium without the editor reaching the server. ### Improved * Faster Large History Scans - Search and file-change discovery now do less unnecessary parsing, making large histories and shell-heavy sessions faster to process. * Lower Memory Use on Big Histories - Session caches are now kept bounded so long-running use and very large histories stay more responsive. * Shell Edit File Cards - File cards now better explain changes made by shell commands, including clearer badges, notes, and command context when an exact diff cannot be reconstructed. * More Accurate Shell-Generated Diffs - More shell-based file writes and edits are detected and shown in the conversation instead of being hidden as plain terminal output. ## What changed in 0.6.8? Version 0.6.8 was released on 2026-08-16. ### Improved * More Reliable History Search - History search now consistently searches the actual session files, so fresh installs and rebuilt indexes no longer miss message results. * Cleaner Onboarding Experience - The welcome flow is simpler, more native to VS Code, localized, and works better when no sessions are found yet. ### Fixed * Search Box Consistency - Fixed cases where clicking a recent search or searching immediately after typing could jump back to an older query. * Claude Session Discovery - Fixed searches on machines where the writable index folder is separate from the real Claude history folder. ## What changed in 0.6.7? Version 0.6.7 was released on 2026-08-13. ### Added * Advanced Deep Search Options - Deep search now has a clearer Advanced panel with labeled options for author scope, match mode, tool output, and subagent transcripts. * Tool Output and Subagent Search - Deep search can now include tool inputs/results and subagent transcripts when you need to find details that are not in the main chat messages. * Priority Section Control - You can now hide the Priority section from the history list. ### Improved * Clearer Deep Search Feedback - Search results now show how many files were scanned and surface scan failures instead of looking like a clean “no results” response. * Better Priority Session Updates - Sessions that appear in both Priority and date sections now keep active, unread, title, and live-status indicators in sync across both rows. ### Fixed * Deep Search Fallback Reliability - Fixed deep search on machines where the search index directory falls back to an alternate location, including Windows setups where searches could incorrectly return zero results. * Deep Search Provider Resilience - If one assistant source cannot be scanned, deep search now continues with the remaining available sources instead of failing the whole scan silently. ## What changed in 0.6.6? Version 0.6.6 was released on 2026-08-08. ### Added * Session File Sizes - Session file size is now stored, backfilled for existing history, shown in the history list, and available as a sort option. * Smarter Project Session Loading - Project groups can load more sessions on demand, with clearer progress when a project has more history than is currently shown. ### Improved * Faster Large Histories - History loading, project grouping, search results, and session pagination are more efficient for users with many sessions. * Better Search Pagination - Search and history results now report loaded vs. total sessions more clearly and load additional results more predictably. * More Reliable Project Search - Project-filtered search and file timelines now match project paths more consistently, including paths with dots or encoded workspace names. * Smoother Conversation Scrolling - Long conversation views scroll more predictably, with fewer jumps and less header flicker while reading. * More Stable OpenCode Refresh - OpenCode session refresh and related parser behavior are more reliable during live updates. ## What changed in 0.6.4? Version 0.6.4 was released on 2026-08-04. ### New: ccassist MCP * Local MCP Server - ccassist can expose your local chat history to supported AI clients so they can search sessions, pull context, inspect changes, and continue work from past sessions. Off until you turn it on. * Guided MCP Setup - The MCP page introduces itself on first visit and walks you through turning the server on, registering your CLIs, and installing the skill in one step. Reopen the walkthrough any time from the "?" button. ### Added * Cursor CLI Session Support - Browse Cursor CLI sessions alongside Claude, Codex, OpenCode, Grok, and Copilot, including diffs, search, agents view, analytics, source filters, and live refresh. * Session Tags and Notes - Add tags and notes to sessions, filter history by tags, search with tag:, and optionally include tags/notes in Markdown exports. * Session Read State - History activity badges now remember which sessions you have opened, including across VS Code windows. ### Improved * Cleaner Long Conversations - Completed work can collapse behind a concise “Worked for” summary, making large sessions easier to scan. * Staged Project Session Reveal - Large project groups now reveal sessions progressively with clearer show-more controls, keeping the history list more responsive. * Better Session Conversion and Resume Registration - Converted sessions are handled more reliably when moving between supported assistant formats. * Improved Message Rendering - Conversation text, colors, and dynamic message details render more consistently. ### Fixed * Unread Activity Accuracy - Read/unread indicators are more consistent when sessions update or are opened from another window. ## What changed in 0.6.3? Version 0.6.3 was released on 2026-07-28. ### Added * LaTeX Math Rendering - Conversations now render LaTeX math with KaTeX. Copying rendered math gives you the TeX source back. * Hide Individual Sessions - Hide single sessions from the history list, not just entire projects; unhide them anytime from the Hidden tab. ### Fixed * Sub-Agent Ordering - Sub-agents in the agent thread rail now appear in the order they actually started. ## What changed in 0.6.2? Version 0.6.2 was released on 2026-07-27. ### Improved * Worktree-Aware Project Grouping - Sessions from linked worktrees are now grouped more intelligently with their main project, making related work easier to find in the history list. * Branch and Worktree Labels - History rows and conversation views can now show branch and worktree context so you can tell which checkout a session came from. ### Fixed * Project Matching Edge Cases - Fixed cases where the same project could appear as separate groups because of different encoded paths or worktree locations. ## What changed in 0.6.1? Version 0.6.1 was released on 2026-07-27. ### Added * Skills Tab - Browse installed agent skills from the side panel, search them, include or hide plugin skills, open bundled files, and copy the slash command for a skill. * Agent Memories Tab - View saved agent memories from the side panel so reusable project and personal context is easier to find. * Mermaid Diagram Rendering - Conversations can now render Mermaid diagrams with zoom and export support. * Large Session Partial Loading - Very large sessions can now open by loading their most recent messages first, with clear warnings when earlier messages were skipped. * Search Operators - History search now supports boolean operators (AND, OR, NOT), literal phrases with exact:, and regular expressions with re:pattern or /pattern/. Plain searches behave exactly as before. * Search By Author - me: searches only your prompts and ai: only the assistant's replies, on their own or combined with any other operator. * Search Options Dropdown - A dropdown at the left of the search bar picks the scope (all / your messages / assistant replies) and match mode (exact phrase / regex) without typing any syntax; the placeholder changes to describe what the current selection searches. AND, OR, and NOT can also be inserted from it. The typed operators still work for anyone who prefers them. ### Improved * Better Settings Organization - Extension settings are now grouped into clearer sections, making configuration easier to scan. * Clearer Tool Output in Conversations - Tool activity in conversations is easier to scan, making long assistant sessions more readable. * Sharper History Search Results - Search matching now handles multi-term queries more consistently and highlights advanced operators more clearly. * Better Source Filtering - History source filters are clearer and more reliable when switching between assistant sources and new side-panel collections. * Windows Live Status Support - Live status support now works better on Windows instead of being hidden behind platform checks. ### Fixed * Deep Search Result Limits - Deep search limits now apply by session count instead of raw match count, so one noisy session is less likely to crowd out other relevant sessions. * Search Box Typing Stability - Fixed the history search box occasionally replacing what you typed with a stale, lowercased copy while results refreshed. ## What changed in 0.6.0? Version 0.6.0 was released on 2026-07-21. ### Added * Live Session Status - A new real-time status mode can show when supported Claude, Codex, and Grok sessions are working, waiting for permission, or finished without waiting for the history list to refresh. * Live Hooks Controls - You can now enable or disable live status from extension commands or the history view, with clear setup states for each supported assistant. * Live Status Trial and Pro Access - Live status includes a free trial flow, clearer locked states, and upgrade messaging when the trial ends. ### Improved * Faster History Loading - Session lists now appear faster with quicker indexing, lighter session reads, and better handling of large histories. * Smoother Session Activity Indicators - New activity and live-state indicators are easier to notice and keep session rows more accurate while work is happening. * Cleaner Search Onboarding - Search setup and indexing preferences are easier to understand, especially when choosing between faster startup and fuller search. * More Polished Conversation Scrolling - Chat and conversation views now land in the expected position more consistently when opened. ### Fixed * Premium Center and Upgrade Flow Clarity - Account, upgrade, and trial-related messages now behave more consistently across entry points. ## What changed in 0.5.7? Version 0.5.7 was released on 2026-07-14. ### Fixed * Codex Quota Display Accuracy - Fixed cases where Codex quota windows could appear in the wrong slot when usage data included newer window details. ## What changed in 0.5.6? Version 0.5.6 was released on 2026-07-11. ### Added * Grok Usage Tracking - Grok activity now appears in usage views, model breakdowns, quota charts, and status indicators alongside your other assistants. * Session Drilldowns in Usage Analytics - Usage analytics now make it easier to drill into the sessions behind your activity so you can understand where time, tokens, and cost came from. ### Improved * Better Grok History and Search - Grok sessions are parsed more completely, show richer session details, and work more reliably in history, search, conversation views, and file-change summaries. * More Reliable Copilot and Grok File Changes - Copilot and Grok sessions now produce cleaner file-change information across session views and diffs. * More Useful Dashboard Views - Usage charts, plan-window views, hourly/session rollups, and source filters are clearer and more consistent across supported assistants. * Broader Localization Coverage - More dashboard, webview, and extension text is now localized consistently for supported languages. * Smoother Account and Pro Flows - Device linking, subscription checks, upgrade prompts, and related error messages are more reliable and easier to understand. * Lightweight Local Search Storage - Conversation message bodies are no longer saved to the extension's local SQLite search index. ### Fixed * Session Time Accuracy - Usage analytics now handle session timing more consistently when building daily, hourly, and session-level summaries. * Search and Indexing Reliability - Fixed edge cases that could make lightweight search or indexed search less consistent for newer session sources. * Safer Session Deletion - If VS Code cannot move an explicitly deleted session to the system Trash, the operation now stops and leaves the original file untouched instead of falling back to permanent deletion. ## What changed in 0.5.5? Version 0.5.5 was released on 2026-07-02. ### Added * GitHub Copilot Session Support - You can now browse and open GitHub Copilot sessions alongside Claude, Codex, OpenCode, and Grok history. * Fable Usage Tracking - Claude API usage tracking now includes Fable usage and limits where available, including status bar and dashboard visibility. * Weekly Reset History - The dashboard now shows a weekly reset history table so you can see when quota windows reset and how much quota was used or left unused. ### Improved * Faster Multi-Assistant History - Session discovery, parsing, searching, and status bar updates are faster across Claude, Codex, OpenCode, Grok, and Copilot histories. * More Consistent Source Filtering - Search and history filtering now apply source filters more reliably across supported assistants. * Cleaner Multi-Source Internals - Shared source handling has been simplified, making resume, discovery, watching, and metadata behavior more consistent across assistants. ### Fixed * Search Source Filter Accuracy - Fixed cases where session search could ignore the selected assistant/source filter. * Live Session Indicators - Live indicators now handle Copilot sessions more consistently. ## What changed in 0.5.4? Version 0.5.4 was released on 2026-06-23. ### Improved * Faster History and Side Panel Rendering - The side panel now caches more of its repeated UI work, helping history views feel snappier and more responsive. * Better Project Path Caching - Project and session lookup work has been reduced in the side panel, improving performance when browsing larger histories. ## What changed in 0.5.2? Version 0.5.2 was released on 2026-06-22. ### Added * Grok Session Support - You can now browse and work with Grok sessions alongside Claude, Codex, and OpenCode history across the extension. * Model Filter in History - The history view now lets you filter sessions by the models they used, making large histories easier to narrow down. * Pinned Projects - Projects can now be pinned so the workspaces you care about most stay easier to find in the history view. * Project Context Menu - Project headers in the history view now have their own context menu for actions like pinning, hiding, revealing in your file manager, and opening in a new window. ### Improved * More Modern History List - The history view has been refreshed with a cleaner, more minimal look while keeping project grouping and quick actions easy to scan. * Source Logos and Visuals - Claude, Codex, OpenCode, and Grok sessions now have clearer source logos and badges across history, search results, and conversation views. * Localized Status and Dashboard UI - More of the extension, especially status bar indicators and dashboard text, now adapts to the user’s language. * Markdown Sync and Restore Flow - Markdown-based session sync and restore workflows are clearer and better integrated into the history UI. ### Fixed * History Filtering Reliability - Fixed several source and model filtering edge cases so history and search results stay more consistent. ## What changed in 0.5.0? Version 0.5.0 was released on 2026-06-13. ### Added * Session Archive (Pro) - Claude and Codex automatically clean up old session transcripts over time. You can now archive sessions to keep a full-fidelity copy safe from that cleanup, with a dedicated Archived tab to browse them across all your projects. Archived sessions can still be resumed, forked, and copied as commands. Viewing and restoring archived sessions stays free, so you never lose access to your backups. * Markdown Session Sync - Export sessions to a portable Markdown format that you can commit to git and move between devices. Exported sessions can be imported back with their file changes and per-turn summaries intact, or resumed as a native Claude/Codex session - even when the original was created on another computer or operating system. * Markdown Sessions Tab - A new MD tab in the history list shows your exported Markdown sessions, viewable just like regular sessions. * Bulk Session Actions (Pro) - Select multiple sessions at once to export or delete them together, with a single confirmation. * Per-Model Usage Tracking - The status bar and dashboard now break down your weekly API usage by model (Sonnet and Opus), giving you a clearer picture of where your quota is going. * In-App Help - Added help buttons and explanations for the new Markdown and Archived session features so it is clear how each one works. ## What changed in 0.4.81? Version 0.4.81 was released on 2026-06-07. ### Added * Date Format Setting - A new setting lets you choose how dates and times appear throughout the extension: region-neutral ISO or your system regional format. * Resume Destination Preferences - You can now choose where sessions resume by default: terminal, desktop app, or VS Code extension, set separately for Claude and Codex. ## What changed in 0.4.80? Version 0.4.80 was released on 2026-05-31. ### Added * Multi-Agent Workflow View - Conversations that run multi-agent workflows now show them clearly, with an execution timeline, individual agent cards, and final outputs. * Question Prompts in Chat - Conversations now display the questions asked during a session along with the available options, recommendations, and the answer that was chosen. * Workflow Stats in Session Analytics - Session analytics now include a workflow summary showing how many workflows ran, how many agents they spawned, tool activity, and token usage. * Cache Efficiency in Session Analytics - Session analytics now flag cache misses caused by idle gaps. * Day-by-Day Usage Breakdown - The dashboard now includes a day-by-day usage breakdown that fills in automatically in the background. ## What changed in 0.4.72? Version 0.4.72 was released on 2026-05-21. ### Added * Resume in Claude/Codex Desktop Apps - You can now resume supported sessions directly in Claude Desktop or Codex Desktop or their VS Code extensions. * Session Deletion from History - You can now delete sessions directly from the history view and move their files to the system Trash. * Right-Click Session Menu - Session rows now support a context menu for common actions like resume, copy command, rename, pin, copy session ID, and delete. * Live Session Indicators - Active sessions are now easier to spot in history with live indicators for supported sources. ## What changed in 0.4.71? Version 0.4.71 was released on 2026-05-11. ### Fixed * Startup and Database Stability - Fixed several startup-time and database-handling issues. * Session Analytics Accuracy - Fixed an issue where some nested agent costs could be counted incorrectly in session analytics. * Background Indexing Reliability - Fixed race conditions that could affect indexing, session lookups, and related background updates. ## What changed in 0.4.70? Version 0.4.70 was released on 2026-05-10. ### Added * OpenCode Session Support - You can now browse and work with OpenCode sessions alongside Claude and Codex history across the extension. * OpenCode Session Analytics and History Views - OpenCode sessions now appear properly in conversation views, analytics, file history, and search results. ## What changed in 0.4.64? Version 0.4.64 was released on 2026-05-01. ### Fixed * History Refresh Reliability - Fixed a history view caching issue so refreshed or newly indexed sessions appear more reliably without needing extra manual refreshes. ## What changed in 0.4.63? Version 0.4.63 was released on 2026-04-27. ### Added * Claude/Codex Source Filter - Added a source filter in history so you can quickly switch between Claude sessions, Codex sessions, or both. * Sub-Agent Stats in Session Analytics - Session analytics now include more detail about sub-agent activity, including usage and duration. ## What changed in 0.4.62? Version 0.4.62 was released on 2026-04-20. ### Added * Reload Conversation - Added a refresh action in the chat view so you can reload the current conversation without reopening it. * Live Agent Conversation Updates - Agent conversations now update more naturally while work is still in progress. * Search and Review Progress Feedback - Deep search and Review Changes now show clearer progress while results and diffs are being prepared. --- # Claude Code & Codex leaderboard > Developers ranked by API-equivalent cost, output tokens, active hours and streak across Claude Code and Codex, with weekly, 30-day and all-time boards per model family. Opt-in from the free VS Code extension. Source: https://ccassist.dev/leaderboard/ Last updated: 2026-10-02 --- Update to 0.7.6 or later. You join from the Leaderboard tab in the extension, which older versions do not have. Ranked by API-equivalent cost from real local usage, measured by the Claude Code and Codex Assist extension. Track your own usage [How it works](#how-ranking-works) [Install in VS Code](vscode:extension/agsoft.claude-history-viewer)[Install in Cursor](cursor:extension/agsoft.claude-history-viewer) Also: [Antigravity](antigravity-ide:extension/agsoft.claude-history-viewer) · [Trae](trae:extension/agsoft.claude-history-viewer) · [Trae CN](trae-cn:extension/agsoft.claude-history-viewer) · [Open VSX](https://open-vsx.org/extension/agsoft/claude-history-viewer) · [Marketplace](https://marketplace.visualstudio.com/items?itemName=agsoft.claude-history-viewer&utm%5Fsource=leaderboard&utm%5Fmedium=board&utm%5Fcampaign=leaderboard%5Fv1&utm%5Fcontent=header) ## How ranking works * Numbers come from daily usage rollups the extension builds from local Claude Code and Codex session logs. * Cost is API-equivalent: tokens priced at each model's public list price. On a subscription it is what the usage would cost at API rates, not what you paid. * Server-side checks recompute cost from the token counts and hold back implausible days. The top of each board is reviewed by hand, and reviewed profiles carry a verified mark. ## What is shared Per day and per model: token counts, API-equivalent cost, request count, active minutes and turn count. Project names, file paths, prompts and code never leave your machine. You choose to appear under a username, as an anonymous entry, or not at all. [Full details in the privacy policy](/privacy/#leaderboard). ## How to join Install Claude Code and Codex Assist in VS Code (version 0.7.6 or later), open its Leaderboard tab, sign in with your email and pick a username and visibility. Your history is uploaded once, then kept current in the background. Turning sharing off takes you off the board straight away. --- # Pro seats for your whole team > Buying Claude Code and Codex Assist for a team? Volume pricing from 5 seats, one payment and a single invoice. Tell me your seat count and I will send a quote. Source: https://ccassist.dev/teams/ Last updated: 2026-09-09 --- Every seat gets the full Pro extension: unlimited history, diffs and undo on any session, full search, forking and export. One payment covers the team and your finance team gets a single invoice. Tell me how many seats you need and I will send a quote, usually the same day. ## What a team purchase includes * Everything in Pro, for every seat on the team. * One payment for all seats, with a single invoice. * Volume pricing from 5 seats up. * Seats added, removed or moved between people whenever you ask. * Priority email support, direct to me rather than a queue. Seats are set up by hand today, so there is no self-serve team dashboard yet. In practice that means you send the list of emails and I activate them. If your team is big enough that this becomes the bottleneck, say so in the form and it will shape what gets built next. ## Ask for a quote Work email Company (optional) Seats (optional) Anything else? (optional) Send enquiryor email [yash@ccassist.dev](mailto:yash@ccassist.dev?subject=Team%20plan%20enquiry) --- # Privacy Policy > All data processing happens locally on your device, and your chat history never leaves your computer unless you explicitly opt in. Source: https://ccassist.dev/privacy/ Last updated: 2026-09-30 --- Last Updated: September 29, 2026 ## Introduction Claude Code and Codex Assist is an unofficial, community-created VS Code extension that helps you manage your Claude Code and Codex CLI chat history. This privacy policy explains how we handle your data. ## Privacy-First Design Claude Code and Codex Assist is built with privacy as a core principle: * All data processing happens locally on your device * Your chat history never leaves your computer unless you explicitly opt-in to community features * No analytics or tracking is enabled by default * The extension reads from your local `~/.claude/` and `~/.codex/` directories only ## Data Collection ### Default (No Opt-In) When you use Claude Code and Codex Assist without enabling any optional features: * No data is sent to external servers * No analytics or usage statistics are collected * All processing is done locally using WebAssembly (sql.js) * Your chat history remains private on your device ### Optional Community Features (Opt-In Only) Community features such as the public leaderboard only send data after you switch them on. Exactly what the leaderboard sends and shows is described in the [Leaderboard](#leaderboard) section below. **You can disable community features at any time** through the extension. ## Leaderboard The public leaderboard at [ccassist.dev/leaderboard](/leaderboard/) ranks developers by their Claude Code and Codex usage. It is opt-in: nothing is uploaded until you sign in from the Leaderboard tab of the extension, choose a username and visibility, and turn sharing on. ### What is uploaded While sharing is on, the extension uploads daily totals built from your local usage rollup, the same numbers its Analytics tab shows. For each tool (Claude Code or Codex) and each day: * the date * active minutes * turn count * the version of the usage rollup and of the pricing table used to compute cost And for each model used on that day: * the model ID (for example `claude-opus-4-5-20251101`) and reasoning effort * token counts by type: input, output, cache creation (including 1-hour and 5-minute cache writes), cache read and thinking tokens * estimated API-equivalent cost * request count Each upload is tied to your account and to the device it came from (the extension's device ID and the device name, so you can see and remove your devices). Your email address is never shown publicly. **Never sent:** project names, file or folder paths, session IDs, prompts, chat messages, code, languages or tools used, and your login tokens. ### What is public * **Public:** your username, avatar, bio and links, and your stats (API-equivalent cost, token counts, active hours, active days, streaks, ranks and daily history) together with the names of the models you used, on the board and on your profile page at ccassist.dev/leaderboard/u//. * **Anonymous:** your stats are ranked on the board under a label such as "Anonymous #4821". There is no profile page, and your username, avatar, bio and links are not shown. * **Hidden:** nothing is public. You are not ranked, and your data is kept only for your own view. Profile links you add are shown with `rel="nofollow ugc"`. The top of each board is reviewed by hand to keep rankings trustworthy, and uploaded days are checked automatically for plausibility. ### Turning it off and deleting your data * Turning sharing off in the extension stops uploads from that device and removes you from the board immediately. Data already uploaded is kept until you delete it, so you can turn sharing back on later. * Removing a device in the extension deletes that device's uploaded data. * "Delete my leaderboard data" in the extension, or deleting your account, removes your leaderboard profile and all leaderboard rows: daily totals, per-model rows and their revision history. ### Retention Leaderboard data is kept for as long as your leaderboard profile exists, and is deleted with it. ## Error Reporting If you opt-in to anonymous error reporting: * Error messages and stack traces are sent to help us fix bugs * No personal information or chat content is included * Reports are anonymous and cannot be traced back to you * You can disable error reporting at any time ## Third-Party Services Claude Code and Codex Assist may interact with the following services (only if you opt-in): * **Leaderboard API (ccassist.dev):** Stores the daily usage totals you choose to share, as described in the Leaderboard section above * **Error Reporting Service:** Collects crash reports and error logs (anonymous) These services are hosted securely and do not have access to your chat content. ## Data Storage * All local data is stored in an SQLite database within your VS Code extension directory * The database is never uploaded or shared unless you explicitly opt-in * You can delete the database at any time through VS Code's extension settings * Uninstalling the extension removes all local data ## Your Rights You have the right to: * Use the extension completely offline (no internet connection required) * Opt-out of all data sharing features (default state) * Request deletion of any data you've shared (contact us via GitHub) * Export your local database at any time * Review the extension's privacy behavior with a local network proxy ## Cookies and Tracking This landing page and the VS Code extension do not use cookies or tracking scripts. If you enable optional analytics in the extension settings, anonymous usage statistics may be collected, but no personally identifiable information is tracked. ## Children's Privacy Claude Code and Codex Assist is not directed at children under 13\. We do not knowingly collect any information from children. If you believe a child has provided data to us, please contact us via GitHub. ## Changes to This Policy We may update this privacy policy from time to time. Changes will be posted on this page with an updated "Last Updated" date. Significant changes will be announced through the extension's changelog and GitHub repository. ## Contact If you have questions or concerns about this privacy policy or how your data is handled, please: * Open an issue on our GitHub repository: [github.com/yashagldit/Claude-Code-History-VSCode](https://github.com/yashagldit/Claude-Code-History-VSCode) * Contact the developer on Twitter: [@yashagl](https://twitter.com/yashagl) ## Disclaimer **Important:** Claude Code and Codex Assist is an unofficial, community-created extension. It is not affiliated with, endorsed by, or connected to Anthropic, OpenAI, Claude AI, or Codex. This extension is an independent project created to enhance the Claude Code and Codex CLI experience. --- # Terms and Conditions > Terms governing use of Claude Code and Codex Assist, an unofficial, community-created VS Code extension and related website services. Source: https://ccassist.dev/terms/ Last updated: 2026-09-03 --- Last Updated: June 14, 2026 ## Introduction These terms govern your use of Claude Code and Codex Assist, an unofficial, community-created VS Code extension and related website services. ## Unofficial Project Claude Code and Codex Assist is not affiliated with, endorsed by, or connected to Anthropic, OpenAI, Claude AI, Codex, Microsoft, or VS Code. Product names belong to their respective owners. ## Your Data The extension is designed to process your Claude Code and Codex history locally. Optional account, device linking, subscription, error reporting, and statistics features may send limited data to ccassist.dev only when required for those features. See the [Privacy Policy](/privacy/) for details. ## Premium Purchases Premium unlocks paid extension features for the plan shown at checkout. Payments are processed by Polar.sh. You are responsible for using the correct purchase email when buying or restoring access. ## Refunds If Claude Code and Codex Assist does not work for your workflow, email [yash@ccassist.dev](mailto:yash@ccassist.dev) within 7 days of purchase to request a refund. ## Acceptable Use Do not abuse the website, APIs, licensing system, payment flow, or device linking system. Do not attempt to bypass premium checks or interfere with service availability for other users. ## No Warranty The extension and website are provided as-is. They may contain bugs, change over time, or stop working with future changes to third-party tools. To the maximum extent allowed by law, there are no warranties of uninterrupted availability, fitness for a particular purpose, or error-free operation. ## Contact For support, billing, terms, or privacy questions, email [yash@ccassist.dev](mailto:yash@ccassist.dev). --- # Blog: Claude Code and Codex CLI guides > Where Claude Code and Codex CLI keep your sessions, what they cost, how usage limits work and how to undo changes. Practical guides, checked against current versions. Source: https://ccassist.dev/blog/ Last updated: 2026-10-01 --- [![A terminal window beside a wide horizontal meter: small cyan segments fill up one after another inside a long amber bar that is nearly full.](/blog/claude-code-weekly-limit-hero.webp)LatestGuideClaude Code weekly limit and 5-hour limit, explainedHow the Claude Code weekly limit and 5-hour session limit work on Pro and Max, when each resets, how to check what's left, and what to do at the limit.Read the post 1 October 2026 · 10 min read](/blog/claude-code-weekly-limit/) Every post ## All posts All8How-to2Guide2Feature guide2Comparison1Roundup1 No posts of this kind yet. * [![A terminal window beside a wide horizontal meter: small cyan segments fill up one after another inside a long amber bar that is nearly full.](/blog/claude-code-weekly-limit-hero.webp)Guide1 October 2026Claude Code weekly limit and 5-hour limit, explainedHow the Claude Code weekly limit and 5-hour session limit work on Pro and Max, when each resets, how to check what's left, and what to do at the limit.Read ](/blog/claude-code-weekly-limit/) * [![A terminal session timeline with one earlier turn highlighted and an arrow rolling the editor's file back to that point, beside a separate stack of file copies.](/blog/undo-claude-code-changes-hero.webp)How-to1 October 2026How to undo Claude Code changes (rewind, git, recovery)Undo Claude Code changes with /rewind checkpoints, git and VS Code Local History, and recover files a Bash command deleted. Covers Codex CLI too.Read ](/blog/undo-claude-code-changes/) * [![A file tree of session files in a dark editor window, with the older files at the bottom fading away while the newest stay bright.](/blog/where-claude-code-stores-conversations-hero.webp)Guide1 October 2026Where Claude Code stores conversations (and why they vanish)Claude Code saves each session as a JSONL file in ~/.claude/projects and deletes it after 30 days. Exact paths per OS, the file format, and how to keep them.Read ](/blog/where-claude-code-stores-conversations/) * [![An editor window with an AI chat panel on one side and a sidebar of small extension panels: a session list, a diff view and a quota meter.](/blog/best-claude-code-vscode-extensions-2026-hero.webp)Roundup5 July 2026Best Claude Code VS Code extensions in 2026The Claude Code VS Code extensions worth installing in 2026, grouped by job, with install counts and update dates checked on 1 October 2026.Read ](/best-claude-code-vscode-extensions-2026/) * [![A terminal window with a usage table on the left and an editor window with a session list, chart and diff on the right, both reading the same stack of session files.](/blog/ccassist-vs-ccusage-hero.webp)Comparison5 July 2026CCAssist vs ccusage: which Claude Code usage tool?ccusage is a free CLI for Claude Code and Codex usage reports; CCAssist puts usage beside chat history and diffs in VS Code. Tested side by side on one week.Read ](/compare/ccassist-vs-ccusage/) * [![A terminal session on the left feeds stacked token bars into an editor panel on the right that shows a daily usage chart and a quota bar filling toward its limit.](/blog/claude-code-usage-tracker-hero.webp)Feature guide5 July 2026Claude Code usage tracker: tokens, cost per day, quotaHow Claude Code usage is measured: the four token fields in each session file, how cost estimates use API prices, what /usage shows, and a daily cost script.Read ](/claude-code-usage-tracker/) * [![A terminal listing dated session files beside an editor that shows one of those sessions as a readable conversation with a diff.](/blog/codex-cli-history-viewer-hero.webp)Feature guide5 July 2026Codex CLI history: find, resume and read old sessionsWhere Codex CLI keeps session history, how codex resume and codex fork find it, jq one-liners for rollout files, and viewing it all in VS Code.Read ](/codex-cli-history-viewer/) * [![A terminal session list, a stack of session files and an editor side panel all pointing back to one earlier conversation.](/blog/view-claude-code-chat-history-hero.webp)How-to5 July 2026How to view Claude Code chat historyView Claude Code chat history: claude --continue, --resume and the /resume picker, jq one-liners for the JSONL files, and searching every project in VS Code.Read ](/how-to/view-claude-code-chat-history/) --- # Best Claude Code VS Code extensions in 2026 > The Claude Code VS Code extensions worth installing in 2026, grouped by job, with install counts and update dates checked on 1 October 2026. Source: https://ccassist.dev/best-claude-code-vscode-extensions-2026/ Published: 2026-07-05 Last updated: 2026-10-01 --- **TL;DR:** Start with Anthropic's own **Claude Code for VS Code** (`anthropic.claude-code`). It runs Claude in the editor and already handles diffs, checkpoints and resuming past chats. Then add one extension for the job it doesn't do: **Claude Code Usage** (`growthjack.claude-code-usage`) for quota in the status bar, **Codex History Viewer** (`hiztam.codex-history-viewer`) for free tagging and search across Claude Code and Codex, or **CCAssist** (ours, `agsoft.claude-history-viewer`) for diffs, search and cost across several CLIs in one place. All numbers below were checked on 1 October 2026. ## What are the best Claude Code VSCode extensions right now? The best extension depends on the job, so this list is grouped that way. I searched the VS Code Marketplace for "claude code", "claude history", "claude session", "claude code usage", "codex" and "codex history" on 1 October 2026. Then I read each candidate's listing and README. Most of the results are chat front-ends, themes, API switchers or abandoned experiments. The ones below are maintained, clearly about Claude Code (or Claude Code plus Codex), and good at one thing. Disclosure: I build **Claude Code and Codex Assist (CCAssist)**, and this page is on its site. It appears where it fits, and the list names a better pick wherever one exists. ![Table of picks by job: Claude Code for VS Code for running Claude Code, 26,726,580 installs; Codex for running Codex, 14,981,510; Claude Code Usage for quota in the status bar, 64,863; Codex History Viewer for free session organizing, 7,791; CCAssist for diffs, search and cost across CLIs, 11,133; Claude Code Sessions for session housekeeping, 9,822; Claude Code Notifier for alerts, 5,791](https://ccassist.dev/blog/best-claude-code-vscode-extensions-2026.png "Recommended extension per job. VS Code Marketplace installs and last-update dates, read from the Marketplace API on 1 October 2026.") ## Pick by job - **Run Claude Code inside VS Code:** Claude Code for VS Code (`anthropic.claude-code`). - **Run Codex CLI's agent in the same editor:** Codex (`openai.chatgpt`). - **See 5-hour and weekly quota at a glance:** Claude Code Usage (`growthjack.claude-code-usage`). For something smaller, Claudemeter (`hypersec.claudemeter`). - **Organize, tag and search Claude Code and Codex sessions for free:** Codex History Viewer (`hiztam.codex-history-viewer`). - **Rename, move and clean up Claude Code sessions:** Claude Code Sessions (`es6kr.claude-sessions`). - **Read old Claude Code chats with no setup:** Claude Code History (`doorsofperception.claude-code-history`). - **See which session changed a file, with diffs, search and cost across Claude Code, Codex and other CLIs:** CCAssist (`agsoft.claude-history-viewer`, ours). - **Get told when Claude is waiting for you:** Claude Code Notifier (`erdemgiray.claude-code-notifier`). ## Comparison table (1 October 2026) Install counts are VS Code Marketplace installs, read from the Marketplace API on 1 October 2026. "Last update" is the Marketplace's last-updated date. Open VSX counts downloads, a different counter, so it isn't mixed in here. | Extension | ID | Job | Marketplace installs | Last update | |---|---|---|---:|---| | Claude Code for VS Code | `anthropic.claude-code` | Official Claude Code panel | 26,726,580 | 30 Sep 2026 | | Codex – OpenAI's coding agent | `openai.chatgpt` | Official Codex panel | 14,981,510 | 23 Sep 2026 | | Chat for Claude Code | `AndrePimenta.claude-code-chat` | Alternative chat UI | 524,912 | 30 Sep 2026 | | Claude Code Usage | `growthjack.claude-code-usage` | Quota and cost in status bar | 64,863 | 24 Sep 2026 | | Claude Status | `long-kudo.vscode-claude-status` | Quota and cost in status bar | 16,500 | 28 Jul 2026 | | Claude Code Usage Tracker | `YahyaShareef.claude-code-usage-tracker` | Quota in status bar | 14,036 | 19 Sep 2026 | | Claudemeter | `hypersec.claudemeter` | Quota and context in status bar | 4,090 | 13 Aug 2026 | | Claude Code and Codex Assist (ours) | `agsoft.claude-history-viewer` | History, diffs, search, cost | 11,133 | 26 Sep 2026 | | Claude Code Sessions | `es6kr.claude-sessions` | Session housekeeping | 9,822 | 18 Jul 2026 | | Codex History Viewer | `hiztam.codex-history-viewer` | History, tags, search | 7,791 | 28 Sep 2026 | | Claude Code History | `doorsofperception.claude-code-history` | Transcript reader | 3,913 | 25 Sep 2026 | | CC Workflow Studio | `breaking-brake.cc-wf-studio` | Visual agent workflows | 10,122 | 11 Jul 2026 | | Claude Code Notifier | `erdemgiray.claude-code-notifier` | Alerts when Claude waits | 5,791 | 9 Sep 2026 | | Claude Changes | `FotisPanokis.claude-changes` | Review and revert edits | 3,136 | 6 Apr 2026 | ## The official integration ### Claude Code for VS Code (Anthropic) This is the extension to install first. Anthropic's [VS Code docs](https://code.claude.com/docs/en/vs-code) call it the recommended way to use Claude Code in VS Code. It has 26,726,580 Marketplace installs, version 2.1.286 on 30 September 2026, and a 3.72 rating from 809 reviews. **Good for:** a chat panel with the file and selection as context, inline diffs you accept or reject, plan review, and @-mentions. It also has [checkpoints](https://code.claude.com/docs/en/vs-code#rewind-with-checkpoints) that rewind code to any message. The **Session history** button lets you [search and resume past conversations](https://code.claude.com/docs/en/vs-code#resume-past-conversations), and `/usage` shows plan limits. **What it lacks:** it isn't a history browser. Session history lists conversations for resuming, not a cross-project archive with file diffs. It also can't keep old sessions alive: Claude Code deletes transcripts older than [`cleanupPeriodDays`](https://code.claude.com/docs/en/settings-reference#cleanupperioddays), which defaults to 30 days. [Where Claude Code stores conversations](https://ccassist.dev/blog/where-claude-code-stores-conversations/) covers how to keep them longer. Note that the extension bundles its own copy of the CLI. To run `claude` in the integrated terminal, you still need the standalone install. ### Codex (OpenAI) If you also use Codex, OpenAI's `openai.chatgpt` extension (14,981,510 installs, updated 23 September 2026) is the Codex equivalent of Anthropic's panel. It works with ChatGPT Plus, Pro, Business, Edu and Enterprise plans, and it can hand larger tasks to Codex in the cloud. It doesn't browse old Codex CLI sessions in depth. The [Codex CLI history viewer](https://ccassist.dev/codex-cli-history-viewer/) guide covers that. ### Chat for Claude Code (alternative UI) Chat for Claude Code (`AndrePimenta.claude-code-chat`, 524,912 installs, updated 30 September 2026) is a third-party chat interface for the Claude Code CLI. It adds git-based checkpoints, an MCP and skills browser, and cost and token tracking. Pick it if you prefer its interface. Most people are better served by the official panel, which Anthropic maintains alongside the CLI. ## History and search Claude Code writes every conversation to a JSONL file (one JSON object per line) under `~/.claude/projects/`. These extensions read those files so you can find old work. For the manual route, see [how to view Claude Code chat history](https://ccassist.dev/how-to/view-claude-code-chat-history/). ### Codex History Viewer (hiztam) The best free organizer. Despite the name, it reads both Codex CLI and Claude Code sessions. It has 7,791 installs, a 4.88 rating from 8 reviews, and version 2.15.0 from 28 September 2026. It is the most actively developed extension in this group. **Good for:** pins, tags, notes, custom titles and saved searches. Search supports regex and boolean queries. There are file-change diff cards, a per-file "AI change history", history insights, and resume through the official Codex or Claude Code extensions. It is free. **What it lacks:** cost and quota tracking. Its README says live quota, billing and cost data are outside the project's scope. It also reads only Codex and Claude Code, not other CLIs. ### Claude Code Sessions (es6kr) The best for housekeeping. It has 9,822 installs and version 0.4.12 from 18 July 2026. **Good for:** a sidebar tree of every project and session. You can rename, delete or drag sessions between projects, bulk-remove empty ones, split sessions, and resume in a terminal or the official extension. **What it lacks:** Codex support, file diffs and any usage view. It is a file manager for sessions, not an archive. ### Claude Code History (doorsofperception) The simplest reader. It has 3,913 installs and version 0.29.3 from 25 September 2026. **Good for:** press `Cmd+Shift+H` and read any past conversation, with tool calls and thinking blocks. A free chat view continues a conversation in place. Reading every conversation is free, and a $9 one-time Pro adds full-text search, bookmarks, statistics and Markdown export. **What it lacks:** Codex support and file diffs. Full-text search is paid. ### Claude Code and Codex Assist (CCAssist, ours) This is our extension. It has 11,133 Marketplace installs, a 3.77 rating from 13 reviews, and version 0.7.5 from 26 September 2026. Separately, Open VSX shows 121,336 downloads, which is where Cursor and VSCodium users get it. **Good for:** answering "which session changed this file, and what did it change?" A file timeline lists every session that edited the open file, with the prompt and a diff you can undo or reapply. It reads Claude Code, Codex, OpenCode, Grok, GitHub Copilot and Cursor sessions in one list, with tags and notes. Search covers prompts, replies and tool output. A usage dashboard shows estimated cost per model and quota burn-down in the same extension. A Recovery tab finds the saved copies Claude Code, Codex and Grok keep of files they edited, so you can compare or restore one after a bad change (today's copies on Free). **What it lacks:** a generous free tier for diffs and usage. Free lets you browse sessions of any age, but search results older than 3 days are blurred and the dashboard shows 7 days. Diffs and undo cover same-day changes, and the file timeline covers 3 days. Full search results, older usage, archiving beyond 3 sessions, forking and bulk export need Pro ([pricing](https://ccassist.dev/upgrade/)). If you mainly want to tag and search months of sessions for free, Codex History Viewer is the better pick. For a status-bar quota number alone, a dedicated usage extension is lighter. ## Usage and cost Two different numbers get called "usage". **Quota** is how much of your 5-hour and weekly plan limits you have used; see [Claude Code weekly limits explained](https://ccassist.dev/blog/claude-code-weekly-limit/). **Cost** is an estimate computed from token counts and published API prices. Pro and Max subscribers don't pay per token, so treat cost as a measure of volume, not a bill. ### Claude Code Usage (growthjack) The most installed usage extension, with 64,863 installs and version 2.4.0 from 24 September 2026. My July version of this page listed a June 2025 update date. That is no longer true: it is actively maintained and now covers Codex too. **Good for:** a status bar with today's cost, the current session's cost, and 5-hour and weekly quota. It also has a dashboard with token composition, per-model tables, and a "where your tokens go" breakdown by content type. **What it lacks:** no chat history, transcript search or file diffs. It is a usage tool only. ### Claude Status, Claude Code Usage Tracker and Claudemeter - **Claude Status** (`long-kudo.vscode-claude-status`, 16,500 installs, 28 July 2026): cost and quota in the status bar, computed from local files, with configurable prices. - **Claude Code Usage Tracker** (`YahyaShareef.claude-code-usage-tracker`, 14,036 installs, 19 September 2026): quota only. It reads the same usage API Claude Code uses, and respects its rate limit. - **Claudemeter** (`hypersec.claudemeter`, 4,090 installs, 4.83 from 6 reviews, 13 August 2026): session, weekly and context-window use, plus Claude service status. All three are narrow on purpose, and that is the point. Pick one status-bar tool, not three. Tools that poll Claude's usage endpoint share a small per-account budget, so stacking them can get you rate-limited. Our [Claude Code usage tracker guide](https://ccassist.dev/claude-code-usage-tracker/) compares the options in more depth. ## Workflow and other jobs - **Claude Code Notifier** (`erdemgiray.claude-code-notifier`, 5,791 installs, 9 September 2026) plays a sound and shows a system banner when Claude needs permission, asks a question or finishes. It also works over SSH, in WSL and in containers. It is worth installing if you run long tasks and look away. - **CC Workflow Studio** (`breaking-brake.cc-wf-studio`, 10,122 installs, 11 July 2026) lets you draw multi-agent workflows on a canvas, then export them as `.claude/agents/` and `.claude/commands/` files. - **Claude Changes** (`FotisPanokis.claude-changes`, 3,136 installs, 6 April 2026) lists files each session changed and reverts them from Claude Code's checkpoint backups. It hasn't been updated in six months, and it can't see edits made through shell commands. For the full picture, read [how to undo Claude Code changes](https://ccassist.dev/blog/undo-claude-code-changes/). ## What changed since the July version - **Claude Code Usage** went from a stale June 2025 release to active updates, and it added Codex. - **Codex History Viewer** grew from 4,268 installs to 7,791 and added history insights and per-file change history. - **Claude Code Sessions** grew from 6,874 installs to 9,822. - The official panel's session history, checkpoints and `/usage` now cover what several small extensions used to do. That is why it leads this list. - **Claude Chats** (`AlexZanfir.claude-chats`) was left out: 8,138 installs, but no update since 1 November 2025. - The standalone desktop app [jhlee0409/claude-code-history-viewer](https://github.com/jhlee0409/claude-code-history-viewer) (2,214 GitHub stars) is still active. It isn't a VS Code extension, so it isn't ranked here. ## How we checked - **Registry data:** VS Code Marketplace `extensionquery` API, read by extension ID on 1 October 2026 (install count, rating, last-updated date, version). Open VSX downloads came from `open-vsx.org/api` on the same day. The two counters are never added together. - **Features:** each extension's current Marketplace README. Anthropic's features come from the [VS Code docs](https://code.claude.com/docs/en/vs-code) and [settings reference](https://code.claude.com/docs/en/settings-reference#cleanupperioddays), read on 1 October 2026. - **CCAssist:** checked against the published 0.7.5 listing and changelog, and against the extension source for free and Pro limits. - **Excluded:** general AI assistants (Cline, Roo Code, Continue), themes, API-provider switchers, and extensions not updated in the last six months, unless they were in the July list. Install CCAssist from the [VS Code Marketplace](https://marketplace.visualstudio.com/items?itemName=agsoft.claude-history-viewer) and it picks up your existing Claude Code and Codex sessions: a project-grouped history list, today's file changes as diffs, and a usage dashboard, all on the free tier. ## Frequently asked questions ### Do I need a third-party extension if I have the official Claude Code extension? No. Anthropic's Claude Code for VS Code (anthropic.claude-code) covers chatting, inline diffs, checkpoints and resuming past conversations. Third-party extensions add things it doesn't do: browsing history across projects and CLIs, cost and quota tracking outside the chat, and session housekeeping. ### Which extension shows Claude Code quota in the status bar? Claude Code Usage by growthjack (growthjack.claude-code-usage) is the most installed option, with 64,863 Marketplace installs on 1 October 2026. It shows cost estimates and 5-hour and weekly quota use for Claude Code, and token use for Codex. ### Can I see Claude Code conversations older than 30 days? Only if the session files still exist. Claude Code deletes transcripts older than cleanupPeriodDays, which defaults to 30. Raise that setting, or archive sessions with an extension, before they age out. --- # CCAssist vs ccusage: which Claude Code usage tool? > ccusage is a free CLI for Claude Code and Codex usage reports; CCAssist puts usage beside chat history and diffs in VS Code. Tested side by side on one week. Source: https://ccassist.dev/compare/ccassist-vs-ccusage/ Published: 2026-07-05 Last updated: 2026-10-01 --- **TL;DR:** If you want a ccusage alternative because you'd rather see usage inside VS Code, next to the conversation and the file changes that produced it, try CCAssist. If you want numbers in a terminal, a script, CI or the Claude Code status line, keep ccusage: it's free, MIT-licensed and better at that job. The two read the same local session files, so you can run both. On one week of my own Claude Code logs, they agreed to the token on six of seven days. ## Is CCAssist a ccusage alternative? Partly. The tools overlap on one thing: both read the session files that Claude Code and Codex CLI write to your disk and turn them into token counts and estimated cost. Beyond that they do different jobs. [ccusage](https://github.com/ccusage/ccusage) is a command-line report generator. You run `npx ccusage@latest` and get a table of tokens and cost by day, week, month, session or 5-hour block, for 18 coding agents. Every report can print JSON with `--json`. CCAssist ("Claude Code and Codex Assist") is a VS Code extension. Its main job is history: browse and search past sessions from Claude Code, Codex CLI, Grok CLI, OpenCode, GitHub Copilot and Cursor, and see each file change as a diff. Usage charts, cost estimates and a quota readout sit alongside that history. For the history side on its own, see [how to view Claude Code chat history](https://ccassist.dev/how-to/view-claude-code-chat-history/). So the honest answer: CCAssist replaces ccusage only if you look at usage in your editor. If ccusage feeds a script or your status line, nothing here replaces it. ![Decision chart: need usage numbers in a script, CI job or the Claude Code status line points to ccusage; need the chat, diffs or search behind a cost figure points to CCAssist; terminal-only work points to ccusage; yes to Q2 plus Q1 or Q3 means use both](https://ccassist.dev/blog/ccassist-vs-ccusage.png "Three questions pick the tool. ccusage 20.0.26 and CCAssist 0.7.5, checked 1 October 2026.") ## What ccusage does well ccusage is the most widely used tool in this space. The `ccusage` npm package had 144,686 downloads in the week of 22–28 September 2026 ([npm API](https://api.npmjs.org/downloads/point/last-week/ccusage)), and the repository had 18,817 GitHub stars on 1 October 2026. It changed a lot this year, and many "ccusage alternatives" pages still describe the old version: - **Version 20 is Rust inside.** Since [v20.0.0](https://github.com/ccusage/ccusage/releases/tag/v20.0.0) (19 May 2026), `npx ccusage` downloads a native binary, so reports on large histories are fast. The current release is v20.0.26 (27 September 2026). - **It covers 18 agents, not just Claude Code.** Claude Code, Codex, OpenCode, Amp, Droid, Codebuff, Hermes Agent, pi-agent, Goose, OpenClaw, Kilo, Kimi, Qwen, GitHub Copilot CLI, Gemini CLI, Antigravity, Grok Build CLI and ZCode, per the [README](https://github.com/ccusage/ccusage#supported-sources). `ccusage daily` combines every source it detects; `ccusage codex daily` narrows to one. - **Separate packages are gone.** `@ccusage/codex` and `@ccusage/mcp` are deprecated on npm with the note to use `npx ccusage` instead. - **Live monitoring was removed.** `ccusage blocks --live` went away in v18 ([docs](https://ccusage.com/guide/live-monitoring)). `ccusage blocks` still prints past and current 5-hour windows. Here is what you'd run: ```bash npx ccusage@latest # all detected agents, by day npx ccusage@latest claude daily --since 20260924 --until 20260930 npx ccusage@latest codex monthly npx ccusage@latest claude daily --instances # split Claude Code by project npx ccusage@latest blocks # Claude Code 5-hour windows npx ccusage@latest daily --json | jq '.totals.totalCost' ``` The JSON is plain and stable enough to script against. A `claude daily --json` run returns a `daily` array and a `totals` object. Each day has `inputTokens`, `outputTokens`, `cacheCreationTokens`, `cacheReadTokens`, `totalTokens`, `totalCost`, `modelsUsed` and a per-model `modelBreakdowns` list. On my machine (42 Claude Code project folders, about 1,600 session files) a one-week report took about 15 seconds through `npx`. The other strength is the status line. Claude Code can run any command as its [status line](https://code.claude.com/docs/en/statusline), and ccusage ships one: ```json { "statusLine": { "type": "command", "command": "npx -y ccusage statusline", "padding": 0 } } ``` That puts session cost, today's cost, the current 5-hour block and a burn rate under your prompt, per the [statusline guide](https://ccusage.com/guide/statusline). It's still marked Beta. It uses cached prices by default so it stays fast. ## What CCAssist does that ccusage doesn't CCAssist answers a different question: not "how much did I spend?" but "which conversation spent it, and what did it change?" - **The session behind the number.** The dashboard and per-session analytics link cost back to the session, and the session opens as a readable transcript with its tool calls. - **File diffs.** Every edit an agent made shows as a diff against the file's earlier state. ccusage never looks at file changes. - **Search across history.** Find the session where you fixed a bug, across Claude Code, Codex CLI and the other supported CLIs. - **Charts instead of tables.** Day-by-day usage with a per-model and reasoning-effort breakdown, and a Weekly Digest tab that compares the last 7 days with the week before. - **Your real quota.** The status bar reads your plan's 5-hour and weekly usage from your Claude account (and Codex's from your ChatGPT account). ccusage's `blocks` groups local logs into 5-hour windows; it doesn't see the limit your account actually has. Our [weekly limit guide](https://ccassist.dev/blog/claude-code-weekly-limit/) explains the two windows. What it doesn't do matters just as much. CCAssist has no command line, no JSON or CSV export of usage, and no status line for Claude Code's terminal. It runs only where VS Code (or a VS Code–based editor) runs, and it reads 6 CLIs where ccusage reads 18. ## Feature comparison Checked on 1 October 2026 against ccusage 20.0.26 and CCAssist 0.7.5. | | ccusage | CCAssist | |---|---|---| | Where it runs | Terminal (Node + native binary) | VS Code and VS Code–based editors | | Price | Free, MIT licence | Free tier; Pro from $4.99/month or $49.99 lifetime | | Claude Code usage and cost | ✓ | ✓ | | Codex CLI usage and cost | ✓ | ✓ | | Other agents | ✓ 16 more (Gemini CLI, Amp, Goose, Kimi…) | Partial: Grok CLI, OpenCode, GitHub Copilot, Cursor | | Daily / weekly / monthly reports | ✓ tables | ✓ charts (free: last 7 days) | | Per-model breakdown | ✓ `--breakdown`, JSON | ✓ | | Per-project split | ✓ `--instances` | ✓ | | JSON output for scripts | ✓ `--json` | ✗ | | Claude Code status line | ✓ `ccusage statusline` (Beta) | ✗ (VS Code status bar instead) | | 5-hour windows | Partial: estimated from local logs | ✓ read from your account, plus weekly | | Custom prices / offline prices | ✓ `ccusage.json`, `--offline` | ✗ | | Browse and read past sessions | ✗ | ✓ (free, any age) | | File diffs of agent edits | ✗ | ✓ (free: same day) | | Search across sessions | ✗ | ✓ (free: full results for 3 days) | | Resume, fork, export sessions | ✗ | ✓ Markdown export (3 free), fork is Pro | | Source code | Open (GitHub) | Closed | ## Why ccusage and CCAssist numbers can differ People comparing tools usually ask this next, so I ran both on the same machine: Claude Code usage for 24–30 September 2026, grouped by local day. | Day | ccusage | CCAssist | Tokens | |---|---:|---:|---| | 24 Sep | $168.72 | $168.72 | identical | | 25 Sep | $72.97 | $72.97 | identical | | 26 Sep | $67.69 | $67.69 | identical | | 27 Sep | $93.56 | $93.56 | identical | | 28 Sep | $160.60 | $170.38 | CCAssist +28.1M | | 29 Sep | $296.12 | $296.12 | identical | | 30 Sep | $77.90 | $77.90 | identical | | **Week** | **$937.56** | **$947.34** | | These are estimates computed from token counts and published API prices. On a Pro or Max subscription you don't pay per token, so treat them as "what this would have cost on the API". Six days match to the token. That's not luck: both tools remove duplicate lines the same way. Claude Code often writes one streamed response as several JSONL lines (JSONL: one JSON object per line) that share a `message.id` and `requestId`. Both tools keep one line per pair, the one with the most tokens. CCAssist modelled this on ccusage, and its 0.7.1 release fixed a case where streamed messages were counted twice. Here is what does cause gaps, with the cause I confirmed for each: 1. **A response copied into a second session file.** The 28 September gap came from one session that was continued under a new session ID: its file started with copies of 148 responses already in the earlier session's file. ccusage removes duplicates across all sessions, since, as its [Claude adapter notes](https://github.com/ccusage/ccusage/blob/main/rust/adapters/claude/src/README.md) say, "Claude Code can copy one response into multiple session transcripts". CCAssist removes them within each session file, so it counted those 148 responses twice. Here ccusage's total is the right one. 2. **A model the price table doesn't know yet.** For Codex CLI on 30 September, both tools counted the same 22.8M tokens, but CCAssist showed $0.02 against ccusage's $4.24. Most of that day ran on a newer model that the extension's price table didn't list yet, so it was priced at $0. ccusage takes prices from [LiteLLM's model list](https://github.com/BerriAI/litellm), fetched at run time unless you pass `--offline`. It prints a "Missing pricing" warning when a model isn't listed. CCAssist uses a built-in table that it refreshes from ccassist.dev. 3. **`/btw` side questions.** Claude Code's [side questions](https://code.claude.com/docs/en/interactive-mode#side-questions-with-btw) can write a side log that repeats the parent's messages. ccusage drops those repeats (the fix for [issue #913](https://github.com/ccusage/ccusage/issues/913)). 4. **Time zones.** Both tools grouped by my local time, which is why the days line up. Pass `--timezone UTC` to ccusage if you compare reports from machines in different zones. Across Codex CLI for the same week, 4 of 6 active days matched exactly. One was the price-table gap above; the other was a single resumed session that Codex later archived, which the two tools placed differently. Neither tool is your bill. Claude Code's own session cost (`cost.total_cost_usd` in the status line data) is also computed on your machine at list price, and [the docs say](https://code.claude.com/docs/en/statusline) it may differ from what you're actually charged. For more on reading these numbers, see our [Claude Code usage tracker guide](https://ccassist.dev/claude-code-usage-tracker/). ## Which one should you use? Go down the list and stop at the first line that fits: - **You want a number in a script, cron job, CI log or dashboard of your own:** ccusage with `--json`. - **You want cost or the 5-hour block under your Claude Code prompt:** `ccusage statusline`. For the real limit percentages without any tool, Claude Code passes `rate_limits.five_hour.used_percentage` and `rate_limits.seven_day.used_percentage` to status line scripts (Pro and Max plans). - **You work over SSH or never open an editor:** ccusage. - **You use Gemini CLI, Amp, Goose, Kimi or another agent CCAssist doesn't read:** ccusage. - **You want to know which session burned the tokens, read it, and see what it changed:** CCAssist. - **You want charts, a weekly summary, or your real 5-hour and weekly quota in the editor's status bar:** CCAssist. - **You want to find or restore work from an old session:** CCAssist (and see [where Claude Code stores conversations](https://ccassist.dev/blog/where-claude-code-stores-conversations/) for what's on disk). - **Several of these:** use both. Neither writes to the session files, so they don't interfere. ## Where CCAssist fits CCAssist is for the moment after ccusage tells you a day cost $170 and you want to know why. It opens that day's sessions, shows the transcript and the files each one changed, and links the tokens back to the model and session that used them. On the free tier you can browse sessions of any age, and you get 7 days of usage charts, same-day diffs, full search results for the last 3 days, and the live quota readout. Pro (from $4.99 a month or $49.99 lifetime, see [pricing](https://ccassist.dev/upgrade/)) removes the time limits and adds session fork, conversion between Claude Code and Codex, and bulk export. It doesn't replace ccusage for scripting, its status line or its wider agent list. Also note the gap above: until the extension removes duplicate responses across session files, a continued session can push its daily total above ccusage's. ## How we compared - ccusage 20.0.26, run with `npx ccusage@latest` on macOS on 1 October 2026; help text, `claude daily --json`, `codex daily --json` and `--instances` output inspected. Duplicate-handling and pricing behaviour read from the [source](https://github.com/ccusage/ccusage/tree/main/rust/adapters/claude). - CCAssist 0.7.5 (Marketplace release of 26 September 2026); features and free vs Pro limits checked in the extension source. - The cost tables compare the two tools' daily totals for the same local session files on one machine. Only counts and totals were compared; no session content is shown. - Download and star counts are from npm and GitHub on the dates stated. Install [CCAssist from the VS Code Marketplace](https://marketplace.visualstudio.com/items?itemName=agsoft.claude-history-viewer) and the side panel lists your Claude Code and Codex sessions from the last week, with usage charts and the file changes behind each one. ## Frequently asked questions ### Is CCAssist a replacement for ccusage? Only if you want usage inside VS Code next to your chat history and file diffs. For JSON output, scripts, CI and the Claude Code status line, ccusage is the better tool, and the two can run side by side. ### Why do ccusage and CCAssist show different costs for the same day? Both remove duplicate streamed lines the same way, so most days match exactly. Gaps come from responses copied into a second session file (ccusage counts them once, CCAssist once per file) and from price tables that don't know a new model yet. ### Does ccusage show my real Claude weekly limit? No. ccusage blocks groups your local logs into 5-hour windows and estimates cost. The real 5-hour and weekly percentages come from your account; Claude Code passes them to status line scripts, and CCAssist reads them for its status bar. ### Is ccusage free? Yes. ccusage is MIT-licensed open source and runs with npx ccusage@latest. CCAssist is free for browsing and searching sessions, with 7 days of usage charts and same-day diffs; Pro starts at $4.99 a month. --- # Claude Code usage tracker: tokens, cost per day, quota > How Claude Code usage is measured: the four token fields in each session file, how cost estimates use API prices, what /usage shows, and a daily cost script. Source: https://ccassist.dev/claude-code-usage-tracker/ Published: 2026-07-05 Last updated: 2026-10-01 --- **TL;DR:** Claude Code records usage in every session file. Each response carries four token counts: input, output, cache write and cache read. A usage tracker drops the repeated copies of each response, prices the tokens at Anthropic's published API rates and groups them by day and model. For today's session, run `/usage` (`/cost` and `/stats` open the same screen). For a daily history, run the script below. If you want a dashboard with charts and quota tracking in VS Code, CCAssist builds one from those same files. ## What a Claude Code usage tracker actually counts Every Claude Code usage tracker, including `/usage` itself, reads the same data. Claude Code saves each session as a JSONL file (one JSON object per line) under `~/.claude/projects//`. The [session files post](https://ccassist.dev/blog/where-claude-code-stores-conversations/) covers the folder layout and the 30-day cleanup. That cleanup matters here: once a file is deleted, its usage history is gone too. Each assistant response line has a `message.usage` object. Here is the shape of one from Claude Code 2.1.286 on 1 October 2026, trimmed to the fields that matter: ```json { "type": "assistant", "requestId": "req_…", "timestamp": "2026-10-01T…Z", "message": { "id": "msg_…", "model": "claude-opus-5-5", "usage": { "input_tokens": 2, "cache_creation_input_tokens": 26409, "cache_read_input_tokens": 26451, "output_tokens": 309, "output_tokens_details": { "thinking_tokens": 49 }, "cache_creation": { "ephemeral_1h_input_tokens": 26409, "ephemeral_5m_input_tokens": 0 }, "service_tier": "standard" } } } ``` The four token types are priced differently: - **`input_tokens`**: new input the model hasn't seen before. In a running session this is often tiny, because almost everything else comes from the cache. - **`cache_creation_input_tokens`**: input written to the prompt cache (Anthropic's store of recently sent text, which makes resending it cheaper). The `cache_creation` object splits it into 5-minute and 1-hour writes, and the two are priced differently. - **`cache_read_input_tokens`**: input served from the cache. Claude Code resends the whole conversation on every request, so this number grows with the length of the session. - **`output_tokens`**: what the model wrote, including thinking. The `thinking_tokens` detail is already part of this number, so don't add it again. ## Why adding up every usage line overcounts Claude Code writes one line per content block (a thinking block, a text block, each tool call). Every one of those lines carries the full `usage` object for the response. In the session above, 49 lines had a `usage` block, but only 19 distinct responses were behind them. Across all my sessions from the 7 days to 1 October 2026, there were 21,700 usage lines but 10,515 distinct responses. Adding every line would have roughly doubled the total. The fix is to keep one copy per `message.id` + `requestId` pair. [ccusage](https://ccusage.com/) and CCAssist both do this. ## How token counts turn into a cost estimate Multiply each token type by its rate, then add them up. These are Anthropic's [published API prices](https://platform.claude.com/docs/en/about-claude/pricing#model-pricing) per million tokens (MTok), checked 1 October 2026: | Model | Input | 5m cache write | 1h cache write | Cache read | Output | |---|---:|---:|---:|---:|---:| | Claude Opus 5.5 | $4 | $5 | $8 | $0.20 | $20 | | Claude Sonnet 5.5 | $2 | $2.50 | $4 | $0.20 | $10 | | Claude Haiku 4.5 | $1 | $1.25 | $2 | $0.10 | $5 | | Claude Fable 5.1 | $10 | $12.50 | $20 | $0.25 | $50 | Cache writes cost 1.25x the input price for 5 minutes and 2x for 1 hour. Cache reads usually cost 0.1x, but Opus 5.5 reads cost 0.05x and Fable 5.1 reads cost 0.025x. A tracker with one flat cache-read rate gets those two models wrong. ![One Claude Code usage block with 2 input, 26,409 one-hour cache write, 26,451 cache read and 309 output tokens, priced at Opus 5.5 list rates to $0.223, mostly from the cache write, then summed per local day](https://ccassist.dev/blog/claude-code-usage-tracker.png "Counts from one real turn in Claude Code 2.1.286; prices from Anthropic's pricing page, checked 1 October 2026. An estimate, not a bill.") The turn in the diagram comes to about $0.22, and 95% of that is the 1-hour cache write. Priced at the 5-minute rate, the same turn reads $0.14. That gap matters because the [Claude Code costs docs](https://code.claude.com/docs/en/costs#why-usage-climbs-in-a-long-session) say the cache lasts an hour on a subscription, so many of your writes land in the pricier bucket. A tracker that ignores the `cache_creation` split undercounts. Token totals and dollar totals also tell different stories. In my own 7 days of data, cache reads were 97.6% of all tokens but 46% of the estimated cost. Output was 0.4% of tokens and 21% of the cost. A token chart shows how much context you resend. The cost chart shows what the expensive work is. These are estimates at list price. They leave out web search fees ($10 per 1,000 searches, which the session file counts but trackers usually don't price) and any negotiated discount. For API billing, the [Console usage page](https://platform.claude.com/usage) is the source of truth. ## Check this session with /usage (and /cost, /stats) Current Claude Code has one usage screen. According to the [commands reference](https://code.claude.com/docs/en/commands), `/cost` is an alias for `/usage`, and `/stats` opens it on the Stats tab. It shows: - **Session block**: tokens by model and a dollar figure, computed locally at list price. It resets on `/clear` (since v2.1.211). - **Prompt cache line**: the share of input served from cache, and cache misses (v2.1.251 or later). - **Plan usage bars** for Pro, Max, Team and Enterprise, plus a breakdown of what used your allowance (skills, subagents, MCP servers). Press `d` or `w` for the last 24 hours or 7 days. The docs name its limits: the breakdown only covers "local session history on this machine", and for subscribers the session dollar figure "isn't relevant for billing purposes". It also doesn't keep a history you can scroll back through day by day. That's the gap a tracker fills. ## Get Claude Code cost per day with a script This Python script reads every session file, drops the repeated copies, prices each response with the table above and prints the last 7 days by local date. It needs Python 3.9+ and nothing else: ```python import json, glob, os from collections import defaultdict from datetime import datetime # USD per million tokens: input, 5m write, 1h write, cache read, output # Anthropic list prices, checked 1 Oct 2026 PRICES = { "claude-opus-5-5": (4, 5, 8, 0.20, 20), "claude-sonnet-5-5": (2, 2.50, 4, 0.20, 10), "claude-haiku-4-5-20251001": (1, 1.25, 2, 0.10, 5), "claude-fable-5-1": (10, 12.50, 20, 0.25, 50), } seen = {} for path in glob.glob(os.path.expanduser("~/.claude/projects/**/*.jsonl"), recursive=True): for line in open(path, errors="ignore"): if '"usage"' not in line: continue try: d = json.loads(line) except ValueError: continue m = d.get("message") if d.get("type") != "assistant" or not isinstance(m, dict) or not m.get("usage"): continue key = (m.get("id"), d.get("requestId")) seen.setdefault(key, (d["timestamp"], m["model"], m["usage"])) # first copy wins days = defaultdict(float) for ts, model, u in seen.values(): p = PRICES.get(model) if not p: continue split = u.get("cache_creation") or {} w1 = split.get("ephemeral_1h_input_tokens", 0) w5 = u.get("cache_creation_input_tokens", 0) - w1 day = datetime.fromisoformat(ts.replace("Z", "+00:00")).astimezone().date() days[day] += (u.get("input_tokens", 0) * p[0] + w5 * p[1] + w1 * p[2] + u.get("cache_read_input_tokens", 0) * p[3] + u.get("output_tokens", 0) * p[4]) / 1e6 for day in sorted(days)[-7:]: print(day, f"${days[day]:.2f}") ``` Three things to know before you trust the output: 1. **Models not in `PRICES` are skipped.** Add a row for every model you use. Run `grep -rhoE '"model":"claude-[^"]+"' ~/.claude/projects | sort | uniq -c` to list them. 2. **Subagent sessions are included**, because the glob searches every subfolder. Subagents use your allowance too. 3. **Only this machine is counted.** Sessions from a laptop, a remote container or claude.ai aren't in these files. For a maintained terminal version with JSON output and per-session reports, use ccusage. The [ccassist vs ccusage comparison](https://ccassist.dev/compare/ccassist-vs-ccusage/) covers when the CLI is the better pick. ## Track usage against a Pro or Max subscription On a subscription you don't pay per token, so the dollar figure means "what this would have cost on the API". It's useful for comparing days or deciding whether a plan pays off. It doesn't tell you when you'll be cut off. What limits you is the percentage used in the rolling 5-hour window and the weekly window. Claude Code shows it in `/usage` and at [claude.ai/settings/usage](https://claude.ai/settings/usage). Those percentages come from Anthropic's servers, not from session files, so a tracker can only show them by asking the same account endpoint. How the windows reset, and how to check what you have left, is covered in [Claude Code weekly limits explained](https://ccassist.dev/blog/claude-code-weekly-limit/). | Question | Where to look | |---|---| | What did this session cost at list price? | `/usage` (or `/cost`) | | What did I use each day this month? | The script above, ccusage or CCAssist | | How much of my weekly allowance is left? | `/usage` plan bars, claude.ai settings, or a quota tracker | | What will my API invoice say? | [Console usage page](https://platform.claude.com/usage) | | Which session burned my 5-hour window? | `/usage` breakdown (by skill, subagent, MCP), or CCAssist plan windows | ## Where CCAssist fits CCAssist (Claude Code and Codex Assist) is a VS Code extension that builds the tracker from your local session files. As of version 0.7.5 (Marketplace, 26 September 2026), running **Open Analytics Dashboard** from the Command Palette shows: - **Day-by-day table** of tokens and cost by local date. You can expand any day into models and reasoning effort, with thinking tokens shown as part of output. - **One view for several CLIs**: Claude Code, Codex CLI, Grok CLI and OpenCode, labelled by source. Codex records usage differently (`token_count` events with `cached_input_tokens`), and the extension reads that format too. See the [Codex CLI history viewer](https://ccassist.dev/codex-cli-history-viewer/) for the Codex side. - **Model trends, per-project totals and cache hit ratio**, with heavy sessions flagged when most of their input missed the cache. - **Quota burn-down**: the 5-hour and weekly percentages for Claude (overall, and per model where your plan has a separate limit) and Codex, with a burn rate and a forecast of whether you'll run out before the reset. There's also a status bar indicator and a history of weekly resets. - **Plan windows**: pick a past 5-hour or weekly window and see which sessions used it, with each session's share. Costs use the same method as the script: one copy per response, the 5-minute and 1-hour split, and published API rates. The extension updates its rate table from ccassist.dev and falls back to a built-in table. **Free vs Pro.** Free shows the last 7 days in the dashboard. It includes the current and previous plan windows, with session detail for recent days. Pro ($4.99 a month or $49.99 lifetime, see [/upgrade/](https://ccassist.dev/upgrade/)) extends the dashboard to 365 days and unlocks older plan windows and per-session detail for any past day or window. **What it doesn't do.** It isn't a bill. It doesn't see usage from other machines or claude.ai. Quota percentages need you signed in to Claude Code or Codex on the same machine, because the extension reads the CLI's saved login to query the account usage endpoint. That endpoint is rate-limited, so the extension checks it every few minutes, not live. ## How we checked - **Session file shape**: Claude Code 2.1.286 on macOS, read 1 October 2026. Field names and counts only. No prompt or project content is shown. - **Deduplication figures**: every assistant response with a timestamp in the 7 days to 1 October 2026 on the author's machine, keyed on `message.id` + `requestId`. - **Token and cost shares**: the same responses, priced at the list rates above (Opus 5.5 was 90% of responses). - **Prices**: [Anthropic pricing page](https://platform.claude.com/docs/en/about-claude/pricing), read 1 October 2026. - **`/usage` behaviour**: [costs](https://code.claude.com/docs/en/costs) and [commands](https://code.claude.com/docs/en/commands) docs, read 1 October 2026. - **CCAssist features and Free/Pro limits**: extension source at the 0.7.5 release (Marketplace, published 26 September 2026). Install [CCAssist from the VS Code Marketplace](https://marketplace.visualstudio.com/items?itemName=agsoft.claude-history-viewer) and run **Open Analytics Dashboard** to see your last 7 days of Claude Code and Codex usage by day, model and project, next to your current quota. ## Frequently asked questions ### Where does Claude Code record token usage? Every assistant response in a session file under ~/.claude/projects/ carries a message.usage object with input_tokens, cache_creation_input_tokens, cache_read_input_tokens and output_tokens. Cache writes are also split into 5-minute and 1-hour buckets. ### Does /cost still exist in Claude Code? Yes, as an alias. In current Claude Code, /cost and /stats both open /usage, which shows the session cost estimate, plan usage bars for subscribers, and a 24-hour or 7-day breakdown. ### Why is my token total so much higher than my bill? Most Claude Code tokens are cache reads, which cost a tenth of the input price or less. Trackers that also add up every line in the session file count some responses two or more times, because Claude Code writes one line per content block with the same usage attached. ### Is the cost figure what I pay on a Pro or Max plan? No. Subscription plans don't bill per token. The dollar figure is what the same tokens would cost at API list prices. On a subscription, the number that limits you is the percentage of your 5-hour and weekly allowance. --- # Claude Code weekly limit and 5-hour limit, explained > How the Claude Code weekly limit and 5-hour session limit work on Pro and Max, when each resets, how to check what's left, and what to do at the limit. Source: https://ccassist.dev/blog/claude-code-weekly-limit/ Published: 2026-10-01 Last updated: 2026-10-01 --- **TL;DR:** Pro and Max plans meter Claude Code two ways at once: a session limit that resets every five hours, and a weekly limit across all models that resets at a fixed day and time assigned to your account. Both are shared with claude.ai and Claude Desktop. Run `/usage` in Claude Code, or open [claude.ai/settings/usage](https://claude.ai/settings/usage), to see how much is left and when each limit resets. When you hit one, you can wait, let Claude Code continue automatically after the reset, or turn on usage credits billed at API rates. ## How the Claude Code weekly limit and 5-hour limit work Anthropic's plan pages describe two meters running at the same time (both updated 22 September 2026): - **Session limit.** "Your session-based usage limit will reset every five hours." The [Max plan page](https://support.claude.com/en/articles/11049741-what-is-the-max-plan) defines Max 5x and Max 20x as five and twenty times the [Pro plan's](https://support.claude.com/en/articles/8325606-what-is-the-pro-plan) per-session allowance. - **Weekly limit.** Pro and Max "also have a weekly usage limit that applies across all models." It "resets at a fixed time each week that is assigned to your account," and that time stays the same no matter when you start using Claude or when your subscription began. Every request counts against both meters. Claude Code's [error reference](https://code.claude.com/docs/en/errors#youve-hit-your-session-limit) warns that "a single burst of heavy activity, such as a large workflow fanout, can exhaust the weekly allowance before the session window resets." ![Illustrative week: 5-hour session windows fill up on days 1 to 5 while weekly usage climbs to the weekly cap on day 5, then usage is blocked until the next weekly reset at the same day and time](https://ccassist.dev/blog/claude-code-weekly-limit.png "Illustrative: window sizes and percentages are invented. The rules (5-hour reset, fixed weekly reset, cap across all models) are from Anthropic's help center, checked 1 October 2026.") The limits are also shared across products. "Both Pro and Max plans offer usage limits that are shared across Claude and Claude Code," says the [Claude Code plan article](https://support.claude.com/en/articles/11145838-use-claude-code-with-your-pro-or-max-plan) (updated 19 August 2026), and that includes Claude Code in VS Code, Cursor and JetBrains IDEs. ### How big is the weekly limit? Anthropic doesn't publish one. The current help pages give no hour counts or message counts. Instead they list what changes how fast you use it: message and context length, tools, model, effort level and multi-step work ([usage limit best practices](https://support.claude.com/en/articles/9797557-usage-limit-best-practices), updated 22 September 2026). Pages that still quote "40–80 hours of Sonnet per week" or a message count per 5 hours are citing figures no current Anthropic page states. The only reliable number is the percentage in your own `/usage` screen. ## Model-specific weekly limits: Opus, Sonnet and Fable On top of the all-model weekly limit, some accounts have a separate weekly cap for one model family: - **Fable.** On Max (and premium Team seats), "you can use up to 50% of your weekly usage limits on Fable models at no extra cost," and Fable uses the limit faster than other models. On Pro and standard Team seats, Fable 5 and 5.1 aren't part of the plan limits at all. They run on usage credits from the first request ([Claude Fable models on your plan](https://support.claude.com/en/articles/15424964-claude-fable-models-on-your-plan), updated 2 September 2026). Settings > Usage shows a separate Fable weekly bar "if included in your plan". - **Opus and Sonnet.** Claude Code still has the messages `You've hit your Opus limit` and `You've hit your Sonnet limit`. The docs say these "each apply only to requests to that model family, so switching to a model outside the family with `/model` keeps you working" ([errors reference](https://code.claude.com/docs/en/errors#youve-hit-your-session-limit), read 1 October 2026). The current help pages don't mention a separate Opus or Sonnet weekly cap for Pro or Max. If your account has one, it appears as its own bar in `/usage`. The practical rule: switching models helps with a model-specific limit. It doesn't help with the session or weekly limit, because those are "shared across all models". ## When do Claude Code limits reset? - **Session limit:** every five hours. The block message names the time, for example `You've hit your session limit · resets 3:45pm`. Settings > Usage shows the time remaining in the current session. - **Weekly limit:** at your account's fixed day and time, for example `You've hit your weekly limit · resets Mon 12:00am`. It doesn't slide forward when you start working. - **Limit resets:** Anthropic occasionally gives eligible plans a free [limit reset](https://support.claude.com/en/articles/17007452-what-is-a-limit-reset) that sets either the 5-hour or the weekly limit back to full. You redeem it from Settings > Usage on the web or in Claude Desktop, not from Claude Code in a terminal or IDE. Your normal weekly reset time stays the same. ### What changed in 2026 Several articles about the weekly limit describe a state that no longer applies: | Date | Change | Source | |---|---|---| | 13 May – 13 Sep 2026 | Claude Code weekly limits 50% higher (promotion). 5-hour limits unchanged. | [Promotion article](https://support.claude.com/en/articles/15910845) | | Jun–Jul 2026 | Community logs showed the "weekly" meter resetting about every 72 hours. Later comments report a return to weekly resets in late July. Never documented by Anthropic. | [GitHub gist](https://gist.github.com/monperrus/3ac4b303a84946bbeaf2b1123ee99491) | | 19 Jul 2026 | Promotion letting Pro and Team use Fable 5 within plan limits ended | [Fable article](https://support.claude.com/en/articles/15424964-claude-fable-models-on-your-plan) | | From 14 Sep 2026 | Claude Code weekly limits permanently 25% higher than before the promotion, for Pro, Max, Team and seat-based Enterprise | [Promotion article](https://support.claude.com/en/articles/15910845) | The 14 September change is why you'll see both "limits up 25%" and "limits cut 17%" headlines. Both are right. Against the pre-May level the limit is 25% higher. Against the promotional level it's 17% lower, because 125 is 17% less than 150. ## How to check how much Claude Code usage you have left Three places show the same server-side percentages: 1. **`/usage` in Claude Code** (`/cost` and `/stats` are aliases). On Pro, Max, Team and Enterprise it shows plan usage bars with reset times, plus a breakdown of what used them: skills, subagents, plugins, MCP servers and scheduled loops. Press `d` or `w` for the last 24 hours or 7 days. The breakdown is "computed from local session history on this machine," so it leaves out claude.ai and other computers ([costs docs](https://code.claude.com/docs/en/costs#plan-usage-breakdown)). If the usage request is rate-limited, `/usage` shows the last bars it loaded in the past hour, with a note, and `r` retries. 2. **[claude.ai/settings/usage](https://claude.ai/settings/usage).** "Current session" shows how much of the five-hour limit you've used and the time remaining. "Weekly limits" shows when the weekly limit resets, for all models and for Fable if your plan has it. Usage credits appear here too. 3. **Your status line.** Claude Code passes `rate_limits.five_hour` and `rate_limits.seven_day` (each with `used_percentage` and `resets_at`) to [custom status line scripts](https://code.claude.com/docs/en/statusline#rate-limit-usage) for Pro and Max subscribers, after the first response in a session. A minimal status line that keeps both numbers on screen: ```bash #!/bin/bash input=$(cat) FIVE_H=$(echo "$input" | jq -r '.rate_limits.five_hour.used_percentage // empty') WEEK=$(echo "$input" | jq -r '.rate_limits.seven_day.used_percentage // empty') [ -n "$FIVE_H" ] && printf '5h: %.0f%% ' "$FIVE_H" [ -n "$WEEK" ] && printf '7d: %.0f%%' "$WEEK" ``` Claude Code also warns before a window runs out, with a line such as `You've used 85% of your session limit · resets 3:45pm`. Keep in mind that the dollar figure at the top of `/usage` is not what you pay on a subscription. It's the same tokens priced at API list rates. How that estimate is built, and how to total it by day, is in the [Claude Code usage tracker guide](https://ccassist.dev/claude-code-usage-tracker/). ## What to do when you hit the limit | Message | What it means | What you can do | |---|---|---| | `You've hit your session limit · resets …` | The 5-hour allowance is used up | Wait (Claude Code can continue automatically), use usage credits, or redeem a limit reset | | `You've hit your weekly limit · resets …` | The weekly all-model allowance is used up | Usage credits, a limit reset, an upgrade, or wait for the weekly reset. Switching models doesn't help | | `You've hit your Opus limit` / `Sonnet limit` | A model-family cap | `/model` to a model outside that family | | Fable consent prompt or "Requires usage credits" in `/model` | Fable is billed to usage credits on your plan, or you've used your Fable share | Confirm to pay with credits, or switch model | The options in more detail: - **Wait and auto-continue.** Since Claude Code v2.1.234, an interactive session signed in with a subscription waits and resumes the interrupted task after the reset, showing `Usage limit reached · continuing automatically at 3:45pm · esc to cancel`. It won't start this wait on its own when the reset is more than 24 hours away, which is common for the weekly limit. Pick "Wait here, then continue automatically" from `/rate-limit-options` in that case ([interactive mode docs](https://code.claude.com/docs/en/interactive-mode#wait-for-a-usage-limit-to-reset)). - **Usage credits.** Run `/usage-credits` (formerly `/extra-usage`) to open Settings > Usage. Credits are prepaid, billed "at standard API rates", can have a monthly spend cap, and have a $2,000 daily redemption limit ([usage credits article](https://support.claude.com/en/articles/12429409-manage-extra-usage-for-paid-claude-plans), updated 24 September 2026). - **An API key.** You can switch to a Claude Console account and pay per token. An `ANTHROPIC_API_KEY` in your environment makes Claude Code bill the API instead of your plan, so check for one if you get charges you didn't expect. - **Upgrade.** Pro to Max 5x, or Max 5x to Max 20x. One caveat on credits: the prompt cache lasts an hour on a subscription but "drops to five minutes once you're drawing on usage credits" ([costs docs](https://code.claude.com/docs/en/costs#why-usage-climbs-in-a-long-session)). Short breaks cost more once you're past the limit. ## Ways to stretch the weekly limit Each of these follows from how Claude Code sends tokens. Every turn resends the whole conversation, your CLAUDE.md and any files Claude has read ([Models, usage, and limits in Claude Code](https://support.claude.com/en/articles/14552983-models-usage-and-limits-in-claude-code), updated 22 September 2026). 1. **Run `/clear` between unrelated tasks.** Anthropic calls it "the single most effective lever for both quality and cost". A session that wandered through three problems pays for all three on every new message. Run `/rename` first so you can find the session again with `/resume` (see [how to view Claude Code chat history](https://ccassist.dev/how-to/view-claude-code-chat-history/)). 2. **Default to Sonnet and save Opus for hard problems.** Opus "uses meaningfully more of your quota". `/model opusplan` plans with Opus and executes with Sonnet. Fable uses the weekly limit faster still. 3. **Don't switch models casually mid-session.** Each model has its own prompt cache, so the first request after a switch re-reads the whole conversation with no cache hits ([errors reference](https://code.claude.com/docs/en/errors#youve-hit-your-session-limit)). 4. **Mind long breaks.** Your first message after more than an hour idle misses the cache and reprocesses your full context. On Pro and Max, Claude Code offers to resume a large session from a summary instead. 5. **Prefer `/clear` to `/compact` when you don't need continuity.** `/compact` reads the whole conversation to summarize it, so on a large context it's a large request. `/clear` costs nothing. 6. **Give subagents a cheaper model.** Subagents keep verbose output out of your main context, but "the subagent's own requests still draw on your usage." Set `model: haiku` for simple ones ([subagent model choice](https://code.claude.com/docs/en/sub-agents#choose-a-model)). Agent teams use about 7x more tokens than a standard session when teammates run in plan mode. 7. **Lower effort for routine work.** Thinking tokens are billed as output, and effort level counts toward usage. Use `/effort` or `/model`. 8. **Watch background work.** Scheduled `/loop` tasks, goal check-ins and messages from your other sessions each send your full context, even while you're away. The `/usage` breakdown lists the heaviest loops. ## Where CCAssist fits CCAssist (Claude Code and Codex Assist) is a VS Code extension. For limits, version 0.7.5 (Marketplace, published 26 September 2026) does three things: - **Status bar quota indicator.** It reads the login Claude Code already saved on your machine (`~/.claude/.credentials.json`, or the macOS Keychain) and asks Anthropic's account usage endpoint for your percentages. The tooltip shows what's left in the 5-hour and weekly windows with reset times. It also shows per-model weekly bars (Sonnet, Opus, Fable) when the account returns them, and usage-credit spend if credits are on. These are Anthropic's numbers, not estimates. The extension refreshes them every 5 minutes while the VS Code window is focused. The same indicator shows Codex CLI's 5-hour and weekly windows from your ChatGPT login, and Grok's weekly credits (more on the Codex side in the [Codex CLI history viewer](https://ccassist.dev/codex-cli-history-viewer/)). - **Burn-down and forecast.** The analytics dashboard (**Open Analytics Dashboard** in the Command Palette) charts each window over time and shows a burn rate with "Exhausts in 1h 20m" or "On pace". **The forecast is an estimate.** It's a straight-line projection from the extension's own snapshots, with idle time counted, so a sudden burst or a quiet afternoon will move it. - **Which sessions used the window.** For a 5-hour or weekly window, the dashboard ranks your local sessions by weighted token use and flags likely causes (long context, cache misses, Opus or Fable, very long sessions). It's a relative ranking built from session files on this machine. It can't see claude.ai or other computers, and Anthropic doesn't publish how it weights tokens against the limit. **Free vs Pro.** The status bar indicator and the burn-down charts are free. Free keeps 7 days in the dashboard and the session breakdown for the current and previous windows. Pro ($4.99 a month or $49.99 lifetime, see [/upgrade/](https://ccassist.dev/upgrade/)) keeps 365 days and opens older windows. **What it doesn't do.** It can't raise or reset your limit, and it isn't a bill. Anthropic rate-limits the usage endpoint, so the numbers can lag by a few minutes. If Anthropic refuses third-party access, the indicator says so and your local stats keep working. For tokens and cost per day, see the [usage tracker guide](https://ccassist.dev/claude-code-usage-tracker/). ## How we checked - **Plan limits and reset rules:** Anthropic help center, read 1 October 2026: [Max plan](https://support.claude.com/en/articles/11049741-what-is-the-max-plan) and [Pro plan](https://support.claude.com/en/articles/8325606-what-is-the-pro-plan) (updated 22 Sep), [Claude Code with Pro or Max](https://support.claude.com/en/articles/11145838-use-claude-code-with-your-pro-or-max-plan) (19 Aug), [usage limit best practices](https://support.claude.com/en/articles/9797557-usage-limit-best-practices) (22 Sep), [usage and length limits](https://support.claude.com/en/articles/11647753-how-do-usage-and-length-limits-work) (16 Sep), [usage credits](https://support.claude.com/en/articles/12429409-manage-extra-usage-for-paid-claude-plans) (24 Sep), [limit resets](https://support.claude.com/en/articles/17007452-what-is-a-limit-reset) (22 Sep), [Fable on your plan](https://support.claude.com/en/articles/15424964-claude-fable-models-on-your-plan) (2 Sep), [weekly limits promotion](https://support.claude.com/en/articles/15910845) (14 Sep). - **Claude Code behaviour:** [costs](https://code.claude.com/docs/en/costs), [commands](https://code.claude.com/docs/en/commands), [errors](https://code.claude.com/docs/en/errors), [interactive mode](https://code.claude.com/docs/en/interactive-mode), [status line](https://code.claude.com/docs/en/statusline) and [model config](https://code.claude.com/docs/en/model-config) docs, read 1 October 2026. Latest Claude Code on npm that day: 2.1.286. - **CCAssist:** extension source at the 0.7.5 release tag (Marketplace, published 26 September 2026). Features built after that release are not described. Install [CCAssist from the VS Code Marketplace](https://marketplace.visualstudio.com/items?itemName=agsoft.claude-history-viewer) to see your Claude Code 5-hour and weekly percentages in the status bar, and a burn-down of each window in the analytics dashboard. ## Frequently asked questions ### When does the Claude Code weekly limit reset? At a fixed day and time that Anthropic assigns to your account. It doesn't move when you start working or when your subscription renews. Settings > Usage on claude.ai and /usage in Claude Code both show the next reset. ### Is the weekly limit shared between claude.ai and Claude Code? Yes. Anthropic's help center says Pro and Max usage limits are shared across Claude and Claude Code, including the IDE extensions and Claude Desktop. ### Does switching to Sonnet help when I hit the weekly limit? No. The session and weekly limits apply across all models. Switching models only helps after a model-specific message such as 'You've hit your Opus limit', or when you reach the Fable share of your weekly limit on Max. ### How many hours of Claude Code do I get per week? Anthropic's current help pages don't give an hour or message count. They describe Max 5x and Max 20x as five and twenty times Pro's per-session allowance, and note that usage depends on model, effort, context length and tools. --- # Codex CLI history: find, resume and read old sessions > Where Codex CLI keeps session history, how codex resume and codex fork find it, jq one-liners for rollout files, and viewing it all in VS Code. Source: https://ccassist.dev/codex-cli-history-viewer/ Published: 2026-07-05 Last updated: 2026-10-01 --- **TL;DR:** Codex CLI saves every session as a JSONL "rollout" file under `~/.codex/sessions/YYYY/MM/DD/`. To continue one, run `codex resume` (current folder), `codex resume --all` (every folder) or `codex resume --last`. To branch without touching the original, use `codex fork`. To read a session without Codex, a few `jq` commands below will list sessions, print your prompts and show which files changed. To browse, search and diff all of it inside VS Code, install CCAssist. ## Where is my Codex CLI history stored? Codex CLI writes one file per session to `~/.codex/sessions/`, grouped into year, month and day folders. The file is a **rollout**: a JSONL file (one JSON object per line) that holds the whole session in order, including your prompts, the model's replies, tool calls, command output and token counts. If you set the `CODEX_HOME` environment variable, everything below lives under that folder instead of `~/.codex`. ![Anatomy of a Codex rollout path: ~/.codex is Codex home, 2026/09/30 is the local date the session started, the rollout name carries the local start time and the thread ID, and a second file with _ continues the same thread. Four cards list session_index.jsonl (thread names), history.jsonl (typed prompts), archived_sessions (archived rollouts) and state_5.sqlite (the thread index the resume picker reads).](https://ccassist.dev/blog/codex-cli-history-viewer.png "Paths from the openai/codex source (rust-v0.159.2) and a real ~/.codex folder, checked 1 October 2026.") Here is what each file in `~/.codex` is for: | Path | What it holds | Useful for | |---|---|---| | `sessions/YYYY/MM/DD/rollout-