claude-replay CLI
A CLI turns coding-agent logs into an interactive HTML replay with tool calls, file diffs and navigable chapters.
- 856 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Exports preserve conversations, sources and tool results in a file that can reveal confidential data.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/claude-replayDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Exports preserve conversations, sources and tool results in a file that can reveal confidential data.
This is third-party code. Review the repository files before installing.
What it does
claude-replay builds a visual replay from a local agent log. It detects the format and maps different clients’ tool calls into a common presentation. The player supports speed changes, collapsible detail and navigation by file. Output is one HTML file without external dependencies. A browser editor and secret redaction support publication preparation. Watch mode presents an ongoing session through a local server.
Who it is for. For developers and technical writers who want to share an agent workflow without a large video recording.
Good fit when
- Attach an agent’s tool sequence to a bug report.
- Embed an interactive code-fix example in documentation.
- Combine several conversations into one demonstration.
- Watch an ongoing session on a remote development machine.
Not a fit when
- You need to restore actual program state from a conversation alone.
- The material contains code the recipient is not allowed to see.
- You need precise timing from a source without timestamps.
Example request
Build a replay of this test session, redact secrets and provide the file for review before publication.Limitations
Node.js 18 or later is required: the CLI processes a saved log and does not record the agent session itself. Export an OpenCode session from its database before using it; otherwise the player cannot read the conversation. Cursor logs have no timestamps, so playback pauses and speed do not reflect actual work time. Inspect the HTML manually after redaction before sharing it: source code, prompts and tool results may still contain information the recipient is not allowed to see.
How to disable. Stop the local server or close the editor. Exported HTML files remain at their chosen destinations.
FAQ
How do I open the editor?
Run `claude-replay` without arguments to open the browser interface and local-session discovery.
Can I pass only a session ID?
Yes. It searches known directories for supported clients. A full path also works.
How do I combine conversations?
Pass multiple identifiers: `claude-replay session1-id session2-id -o combined.html`.
Can the result be embedded in documentation?
Yes. Host the self-contained HTML and embed it through an iframe.
Is this an official Anthropic tool?
No. It is a community project without claimed Anthropic affiliation or endorsement.
Discussion
No questions or reviews yet.
Related
CLI toolsReads Claude Code session logs as a conversation: tool calls, token counts and a live tail of a running session.
A debugging tool for Claude Code that reads the session logs already saved on disk and shows exact file paths, diffs, model thinking and a category-by-category
agentplay turns local Claude Code, Codex and Cursor sessions into a file map and a replayable step timeline.
Three Claude Code plugins: the savvy-flow orchestrator with tiered subagents and two mods for a progress panel and a cache timer
Draws a Claude Code or Codex session as a graph of agents and tool calls, live or replayed from a recording
A Claude Code plugin stores conversations and decisions in Markdown, updates CLAUDE.md and helps resume work in a new session.
Comparison with similar tools
Cards that share the main search query, and how they differ.
README in short
The player includes a file-activity sidebar and bookmarks. A Docker route mounts source-session directories read-only.