save-learning
Skill1 starsChanged 39 days agoMITView on GitHub
Save a short, reusable lesson as a note in docs/learnings/ - the context, the lesson, why it holds, an example and when it applies - citing the plan, roadmap or ticket it came from by id, and list it in the folder's index. Only for insights that are non-obvious and useful beyond the task at hand; says so instead of writing when nothing qualifies. Use when the user says "save this learning", "write down what we learned", "note this lesson" or "add it to the learnings", or accepts an offer to save one after a plan, a debugging session or a research. Not for tasks still to do (new-ticket) or for project rules the agent must always follow.
3 other projects use a version of this.
Adopted by
Most recently taken up by
- athina-ai/goose-skillstheir own version, first seen Nov 9, 2026
- gooseworks-ai/goose-skillstheir own version, first seen Nov 1, 2026
- caliber-ai-org/ai-setuptheir own version, first seen Apr 28, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- athina-ai/goose-skills★ 1.3ksame name
- caliber-ai-org/ai-setup★ 1.3ksame name
- gooseworks-ai/goose-skills★ 1.3ksame name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository12 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- deep-reviewSkill
- find-bugSkill
- implement-planSkill
- make-planSkill
- make-roadmapSkill
- new-ticketSkill
- plan-pageSkill
- plans-conventionSkill
- research-topicSkill
- ship-workSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 3 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 26
Not recorded elsewhere: not already obvious from the code, a comment, the docs or the commit history.
- Distinctive: line 29
If it fails the bar, tell the user why in one line and suggest the right place instead (a code comment, the README, a ticket).
- Very specific: line 38
Read docs/learnings/README.md and search the folder for the topic.
- Very specific: line 44
Create docs/learnings/<slug>.md from templates/learning.md.
- Distinctive: line 56
Write it in English, like every artifact of the toolkit, and verify every fact, path and snippet in it: a wrong learning is worse than none.
- Very specific: line 61
Add one line to docs/learnings/README.md, creating it with a # Learnings title if it does not exist, and keep the list sorted by date, …
- Very specific: line 71
If the lesson comes from a plan, add a line to that plan's *Closure* (or *Implementation*, if it is still open): Learning: …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Save a learning
A learning is something we found out the hard way and would like to know next time: a library's real behaviour, why an approach failed, a trick that made something work. Written down briefly, it teaches whoever reads it, person or agent, in a minute. Written down at length or for everything, the folder becomes a changelog nobody reads, so the bar matters as much as the writing.
Load the plans-convention skill first when the work relates to a plan.
1. Decide whether it deserves a note
Find the lesson in the conversation, the plan, the debugging report or the research the user is referring to. Then check it against the bar:
- Non-obvious: a competent developer would not have known it in advance.
- Reusable: it helps beyond this task: in another feature, project or decision.
- Explained: we know why it holds, not only that it happened once.
- Not recorded elsewhere: not already obvious from the code, a comment, the docs or the commit history.
If it fails the bar, tell the user why in one line and suggest the right
place instead (a code comment, the README, a ticket). If the conversation
holds several lessons, propose the ones that pass, one line each, and let
the user pick. A rule the agent must follow on every task in this project
belongs in the project's agent instructions file (AGENTS.md, CLAUDE.md),
not here; offer that instead.
2. Avoid duplicates
Read docs/learnings/README.md and search the folder for the topic. If a
note already covers it, update that note (add the new case, sharpen the
lesson, add the source) instead of writing a second one.
3. Write the note
Create docs/learnings/<slug>.md from templates/learning.md. The slug is
the lesson in a few words (sqlite-wal-needs-shared-memory), not the task
(plan-0042-notes).
- Title: the lesson as a claim ("SQLite in WAL mode needs shared memory, so it fails on network drives"), not a topic ("SQLite notes").
- Brief: the whole note fits on one screen. Cut the story of how it was found; keep the lesson, the why and the smallest example.
- Sources: cite the plan, roadmap or ticket by id (
0042), never by path, because folders move when their status changes; add the commit, PR or external source when they help. If the lesson does not come from a plan, leavesourceempty. - Write it in English, like every artifact of the toolkit, and verify every fact, path and snippet in it: a wrong learning is worse than none.
4. Update the index
Add one line to docs/learnings/README.md, creating it with a
# Learnings title if it does not exist, and keep the list sorted by date,
newest first:
- [SQLite in WAL mode needs shared memory](sqlite-wal-needs-shared-memory.md) — 2026-10-03 · from 00425. Link back and report
If the lesson comes from a plan, add a line to that plan's Closure (or
Implementation, if it is still open):
Learning: docs/learnings/<slug>.md. Then report the path and the lesson in
one line each.
Red flags
- A note that retells what happened instead of what we learned.
- A lesson without its why.
- Saving something obvious, or something the code already says.
- Citing a plan by path, or a plan id that does not exist.
- A second note on a topic that already has one.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

