agentleFS
Sign inStart free
sschemelsschemel/second-brain-starter/.agents/skills/freddie-learn/SKILL.md

freddie-learn

Skill0 starsChanged 16 days agoMITView on GitHub

Logs a behavior correction, failure mode, or lesson. Global agent-behavior rules go to 00 - Inbox/Learnings.md (folded into AGENTS.md at the next /eod); lessons specific to one project go to that project's Learnings.md. Use when the user types /freddie-learn, corrects how the agent did something (\"don't do it that way\", \"always…\", \"never…\"), edits the agent's output to fix a mistake, or a non-obvious project gotcha is discovered.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If it's unclear, prefer project.

    line 16
  2. Distinctive:

    Read the destination file and check for an existing entry on the same topic.

    line 17
  3. Very specific:

    If invoked automatically mid-task, do it silently and keep working.

    line 25

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Freddie-Learn: Capture a Correction

Correction or rule: $ARGUMENTS (if empty, infer it from the user's most recent correction in this conversation).

  1. Distill it into a failure mode (what went wrong, specifically) and a corrected rule (an imperative that is reusable beyond this one case).
  2. Route it. Pick one destination:
    • Global (how agents should behave anywhere: workflow, tone, tooling habits): 00 - Inbox/Learnings.md. The next /eod (or, if that is missed, the next /sod catch-up) promotes it into AGENTS.md §7.
    • Project (true only for one codebase or project, e.g. "this repo's dev server serves stale CSS; restart with a clean build folder"): 10 - Projects/<P>/Learnings.md. Create the file with a # <Project> Learnings heading if it doesn't exist. These stay in the project and are read when working on it. /eod only copies a general version into AGENTS.md when the lesson holds beyond the project.
    • If it's unclear, prefer project. A project lesson that keeps recurring elsewhere can be promoted later.
  3. Read the destination file and check for an existing entry on the same topic. If one exists, sharpen that entry instead of adding a duplicate.
  4. Append the entry (get the date from date +%F):
    ### 📅 YYYY-MM-DD - <Topic>
- **Failure Mode**: <what happened>
- **Rule**: <the corrected instruction>
  5. Scrub anything secret-shaped (keys, tokens, customer data) before writing.
  6. If invoked automatically mid-task, do it silently and keep working. If invoked by /freddie-learn, confirm in one line where it went (the Inbox for the next /eod sweep, or the project's Learnings.md).

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.

Check your setup, free