
Learn: Save Session Lessons
Distills useful lessons from a chat and stages them as proposed skill-library updates.
- Repository
- tigerless-labs/autoharness
- Stars
- 10,630
- Licence
- MIT
- Files
- 1 file
What it does
When you type /learn or ask the agent to remember something, it reviews the current conversation and picks out reusable lessons, such as corrections you made, useful techniques, or your lasting preferences. It checks whether an existing skill already covers the topic and prefers updating it over creating a duplicate. Changes are not written directly; they are submitted through a stage_skill tool so a separate promoter can validate them, and rejections are reported at the next session start.
Great for
- Developers who maintain a personal or team library of agent skills and want lessons captured as they work
- Users who correct the agent repeatedly and want those corrections kept for future sessions
Skip it if
- One-off tasks or environment-specific failures, which the skill itself says to skip because they are not reusable
- Shared or production skill libraries where automatic proposed edits should be reviewed by a human first, since it rewrites statements across managed skill trees to resolve contradictions
Try asking
- “/learn”
- “Remember that I prefer tests written before implementation in this project”
- “Save the lesson from that debugging session so we don't repeat the same mistake”
- “Distill what we just figured out about the API pagination into the skill library”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
learn.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip learn.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip learn.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Review the staged proposals and rejections reported at session start rather than assuming every lesson was saved
- Give a clear, specific correction in the conversation so the evidence quote is meaningful
What's inside
- SKILL.md
Redistributed unmodified from tigerless-labs/autoharness under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.