agentleFS
Sign inStart free
game1991game1991/claude-skills/skills/learning-from-experience/SKILL.md

learning-from-experience

Skill0 starsChanged 6 months agoMITView on GitHub

Use when user repeats similar requests or says "learn from this" / "save this pattern". Routes experience to SKILL.md, script/MCP proposals, or MEMORY.md rule upgrades. Does NOT duplicate Auto Memory — focuses on turning experience into reusable artifacts.

4 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions4 projects use a version of this.

    Word for word · 1

    1. game1991/awesome-skills★ 0100% same

    Edited · 0

    None found.

    Wrote their own version · 3

    1. FlyFission/nuclear-grade-context-engineering★ 33same name
    2. prime-radiant-inc/study-skills★ 9same name
    3. agrivisionai/claude-skills★ 1same 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.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    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:

    Always check catalog first: No shortcuts.

    line 46
  2. Very specific:

    When 3+ rules under same topic → mark for SKILL.md upgrade proposal

    line 73

≈ “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

Learning from Experience

Core principle: Auto Memory remembers. LFE upgrades — from experience to reusable artifact.

NOT a replacement for Auto Memory. Claude Code already auto-extracts feedback and preferences. LFE adds what Auto Memory cannot: artifact generation (SKILL.md, scripts, MCP proposals) and explicit pattern promotion (rule → skill upgrade).

REQUIRED: Understand superpowers:writing-skills before generating SKILL.md. This skill decides WHAT; writing-skills decides HOW.

When to Use

DO trigger when:

  • User repeats a request that matches a cataloged pattern
  • User says "learn from this", "save this pattern", "make a skill for this"
  • A MEMORY.md rule has been reinforced 3+ times → propose SKILL.md upgrade

DO NOT trigger for:

  • Single-session corrections (Auto Memory handles this)
  • Simple preference recording (Auto Memory handles this)
  • Session-end summarization (AutoDream handles this)

Quick Reference — Artifact Routing

ConditionArtifactAction
3+ MEMORY.md rules on same topicSKILL.mdDelegate to writing-skills
General methodology (>=3 steps)SKILL.mdDelegate to writing-skills
Repeated automatable operationScript/MCP proposalProvide spec, user decides
Simple rule needing explicit user controlMEMORY.md ruleEdit with [LFE-Pxxx] tag

Note: Most simple rules should be left to Auto Memory. Only write MEMORY.md rules manually when you need explicit [LFE-Pxxx] tags for future migration tracking.

Core Pattern

Step 1: Detect Pattern Match

  1. Read ~/.claude/skills/learning-from-experience/references/pattern-catalog.md
  2. Match user request keywords against catalog entries
  3. Hit found → AskUserQuestion: "Detected [Pxxx] pattern. Reuse / New / Improve?"
  4. No hit → Only proceed if user explicitly triggers ("learn from this")

Always check catalog first. No shortcuts.

Step 2: Route to Artifact

Based on Quick Reference table:

  • SKILL.md → Output delegation package (Step 3)
  • Script/MCP → Provide specification, AskUserQuestion to confirm
  • MEMORY.md rule → Edit with [LFE-Pxxx] tag, update catalog (rare — prefer Auto Memory)

Step 3: Delegate to writing-skills

For SKILL.md artifacts, output this delegation package then invoke superpowers:writing-skills:

## Delegation to writing-skills
- **Pattern name**: [skill-name-with-hyphens]
- **Trigger**: [Use when... description]
- **Core process**: [step summary]
- **Known test scenarios**: [pressure scenarios from experience]
- **Key constraints**: [boundary conditions from experience]

Step 4: Update Catalog

Append or update pattern in pattern-catalog.md:

  • New pattern: assign next Pxxx ID, status "observing" (frequency 1) or "confirmed" (frequency >=3)
  • Existing pattern: increment frequency, update last-triggered date, change status if threshold met
  • When 3+ rules under same topic → mark for SKILL.md upgrade proposal

What LFE Does NOT Do (Left to Built-in)

Built-in handlesLFE does not
Auto-extract feedback from correctionsAuto Memory
Auto-consolidate memories across sessionsAutoDream
Record simple user preferencesAuto Memory
Semantic search of past conversationsepisodic-memory plugin

Common Mistakes

MistakeFix
Writing simple rules to MEMORY.mdLeave to Auto Memory — LFE only writes tagged rules for migration tracking
Creating SKILL.md directlyDelegate to writing-skills for TDD quality
Triggering on every session endAutoDream handles consolidation — LFE only triggers on explicit pattern match
Skipping catalog checkAlways scan catalog first — prevents duplicates
No user confirmation before creatingAlways AskUserQuestion before any artifact

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