# Learn

> >

- **Type:** Skill
- **Install:** `agentstack add skill-cookys-autopilot-learn`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [cookys](https://agentstack.voostack.com/s/cookys)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [cookys](https://github.com/cookys)
- **Source:** https://github.com/cookys/autopilot/tree/develop/skills/learn

## Install

```sh
agentstack add skill-cookys-autopilot-learn
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# learn

Record reusable knowledge so future sessions avoid the same mistakes.

## When to Record

| Trigger | Category |
|---------|----------|
| Bash command failed on path/config, then fixed | `env` |
| Compile error took 2+ retries to fix | `build` |
| Searched 3+ times to find a file/class | `env` |
| API usage was wrong, then corrected | `api` |
| Architecture decision required iteration | `arch` |

## Flow

1. **Dedup check** — search existing knowledge before writing:
   ```bash
   grep -ri "" .claude/knowledge/*.md
   ```
   - Already recorded and complete: skip
   - Recorded but incomplete: update existing entry (Edit)
   - Not found: add new entry

2. **Read** the target knowledge file (by category):
   - `build-errors.md` | `debug-patterns.md` | `architecture.md` | `environment.md`

3. **Append** entry using this template:
   ```markdown
   ## [Title]
   **Date**: YYYY-MM-DD | **Context**: what you were doing
   **Problem**: what went wrong
   **Solution**: what fixed it
   ```
   Add `**Failed attempts**:` and `**Related**:` lines only when they add value.

4. **Update INDEX.md** "recent learning" table

5. **Rotate** — keep recent list at 10 entries max:
   ```bash
   # If your project has a rotate script:
   node .claude/scripts/learn-rotate.js
   # Otherwise: manually trim the oldest entries
   ```

## Categories

| Category | File | Content |
|----------|------|---------|
| `build` | `build-errors.md` | Compile/link/CMake errors |
| `debug` | `debug-patterns.md` | Debugging techniques, crash patterns |
| `arch` | `architecture.md` | Design decisions, pitfalls |
| `env` | `environment.md` | Paths, Docker, config |
| `api` | (in relevant file) | API misuse patterns |

## Invocation

```
/learn [category] [brief description]
```

Examples:
```
/learn build Missing dependency — install with apt/brew/pip
/learn env Script path wrong — use relative path from project root
```

---

## Session Learning Summary (L-size)

At the end of L-size tasks, produce a structured summary before the dev-flow session-end phase:

```markdown
### Errors Encountered
- [error]  — Root cause:  — Fix:  — Recorded? Y/N

### Key Decisions
- [decision]  — Reason: 

### Surprises
- 

### Action Items for Future Sessions
- [ ] Record  to knowledge/.md (if not yet recorded)
- [ ] Update skill  (if pattern repeated 3+ times)
- [ ] Refresh MEMORY.md (if applicable)
```

For S-size tasks: skip the full summary. Instead ask: "Did I retry any operation 2+ times?" If yes, record via standard flow above.

## Knowledge Health Audit

Run a structured health audit across all memory/knowledge files. Produce a status report with actionable remediation.

### 1. MEMORY.md Line Count

```bash
# Find the project memory file
find ~/.claude/projects/ -name "MEMORY.md" -path "*/memory/*" 2>/dev/null | head -1 | xargs wc -l
```

| Status | Lines | Why it matters |
|--------|-------|----------------|
| OK | =200 | **Content after line 200 is silently lost** |

**Remediation**: Move detail sections to `memory/` sub-files, keep MEMORY.md as an index with links.

### 2. Knowledge File Sizes

```bash
wc -l .claude/knowledge/*.md 2>/dev/null
```

| Status | Lines | Why it matters |
|--------|-------|----------------|
| OK | 500 | Too expensive to read in full — entries get skipped |

**Remediation**: Archive old entries to `knowledge/archive/`, or split into sub-topic files.

### 3. Verification Staleness

Read first line of each knowledge file for ``.

| Status | Age | Why it matters |
|--------|-----|----------------|
| OK | 60 days or missing | High risk of wrong advice |

**Remediation**: Read the file, confirm each entry still applies, update the `last-verified` date.

### 4. INDEX.md Recent Learning Count

Count rows in the "recent learning" table (excluding header/separator).

| Status | Count |
|--------|-------|
| OK | 10 — rotation needed |

**Remediation**: Trim oldest entries or run rotation script if your project has one.

### 5. Orphan Files

List `.md` files in `.claude/knowledge/` not referenced in INDEX.md.

**Remediation**: Add to INDEX.md, or delete if obsolete.

### Health Report Output Format

```
Memory Health Report (YYYY-MM-DD)

| Status | Item                  | Value     | Threshold |
|--------|-----------------------|-----------|-----------|
| ...    | MEMORY.md             | N lines   | <=200     |
| ...    | (each knowledge file) | N lines   | <=300     |

Stale Verification:
| File | Last Verified | Days Ago | Status |

INDEX.md Recent Learning: N / 10 — STATUS

Orphan Files: (none) or list

Recommended Actions:
1. [CRITICAL] ...
2. [WARNING] ...
```

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [cookys](https://github.com/cookys)
- **Source:** [cookys/autopilot](https://github.com/cookys/autopilot)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-cookys-autopilot-learn
- Seller: https://agentstack.voostack.com/s/cookys
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.