#176176th of 324 in Skills, plugins and prompts
backpass
CLI that proposes evidence-backed edits to AGENTS.md, CLAUDE.md and skills
- Stars
- 1.5k
- License
- MIT
- Last commit
- Oct 2026
- Last release
- Oct 2026
- Language
- JavaScript
16 on 21 Aug 20261,450 on 30 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
backpass reads agent session transcripts from disk for seven harnesses (Claude Code, Codex, Pi, OpenCode, Grok, Cursor CLI, Hermes), locally or over SSH. It distills them into proposed edits to your AGENTS.md or CLAUDE.md and project skills, each backed by verbatim quotes from at least two sessions. Analysis never writes; you accept or reject each edit in `backpass apply`.
Who it is for: Developers who use coding agents and want to tune AGENTS.md and skills
Strengths
- Every add, rewrite or remove edit needs quotes from at least two distinct sessions
- Reads transcript stores directly from disk; no upload, and it has no API keys of its own
- Supports project scope and user scope (--scope user) plus SSH collection from other machines
- Analysis never writes files; changes only land after explicit review in apply
Weaknesses
- Requires Node >= 22.5 and acpx on PATH; default review UI also needs lavish-axi
- macOS and Linux only; Windows remotes are out of scope
- User scope collects only Claude and Codex sessions by default
- Model calls go through an already-authenticated harness, so usage costs are unknown
What it needs
- no GPU
- Needs Node >= 22.5, acpx, lavish-axi (for default review UI), ssh (optional, for remote hosts)
- Models: Claude Code, Codex, Pi, OpenCode, Grok
Quick answers
Is backpass open source?
Yes. Its code is on GitHub under the MIT license.
Also in Skills, plugins and prompts
See all 324| Rank | Project | Score |
|---|---|---|
| 1 | impeccableDesign skill with 24 slash commands for AI coding agents' frontend output | 96 out of 100 |
| 2 | skillsAgent skills for planning, TDD, debugging and code review | 95 out of 100 |
| 3 | eccAgents, skills, hooks and rules bundle for Claude Code, Codex and other harnesses | 95 out of 100 |
| #4 | gstackMarkdown slash-command skills that give Claude Code planning, review, QA and ship roles | 95 out of 100 |
| #5 | cavemanCuts token use in coding agents with terse replies and compressed inputs | 95 out of 100 |