Structured learning system for continuous self-improvement. Use this skill to log learnings, track errors, record feature requests, detect recurring patterns, and extract new skills from repeated workflows. Triggers on: learning, error tracking, pattern detection, skill extraction, retrospective, what went wrong, lessons learned, improve system.
Continuous learning through structured logging, pattern detection, and automatic skill extraction.
.learnings/
├── LEARNINGS.md # What we learned (LRN entries)
├── ERRORS.md # What went wrong (ERR entries)
└── FEATURE_REQUESTS.md # What users wanted (FEAT entries)
Append to .learnings/LEARNINGS.md:
## [LRN-YYYYMMDD-XXX] category
**Logged**: ISO-8601 timestamp
**Priority**: low | medium | high | critical
**Status**: pending | resolved | promoted
**Area**: frontend | backend | infra | gateway | skills | security
### Summary
One-line description of what was learned
### Details
Full context: what happened, what was wrong, what's correct
### Suggested Action
Specific fix or improvement to make
### Metadata
- Source: conversation | error | user_feedback
- Related Files: path/to/file.ext
- Tags: tag1, tag2
---
Append to .learnings/ERRORS.md:
## [ERR-YYYYMMDD-XXX] component_name
**Logged**: ISO-8601 timestamp
**Priority**: high
**Status**: pending | resolved
**Area**: frontend | backend | infra | gateway | skills | security
### Summary
Brief description of what failed
### Error
Actual error message or traceback
### Context
- Command/operation attempted
- Input or parameters used
- Environment details if relevant
### Fix Applied
What resolved the issue (fill in after resolution)
### Metadata
- Reproducible: yes | no | unknown
- Related Files: path/to/file.ext
- See Also: ERR-YYYYMMDD-XXX (if recurring)
---
Append to .learnings/FEATURE_REQUESTS.md:
## [FEAT-YYYYMMDD-XXX] capability_name
**Logged**: ISO-8601 timestamp
**Priority**: medium
**Status**: pending | planned | implemented
**Area**: frontend | backend | infra | gateway | skills | security
### Requested Capability
What the user wanted to do
### User Context
Why they needed it, what problem they're solving
### Complexity Estimate
simple | medium | complex
### Suggested Implementation
How this could be built, what it might extend
### Metadata
- Frequency: first_time | recurring
- Related Features: existing_feature_name
---
Format: {TYPE}-{YYYYMMDD}-{XXX}
When logging something similar to an existing entry:
grep -r "keyword" .learnings/See Also: ERR-YYYYMMDD-XXX in MetadataReview .learnings/ at natural breakpoints (end of session, end of day, end of week):
# Count by status
grep -c "Status: pending" .learnings/*.md
grep -c "Status: resolved" .learnings/*.md
# Find high-priority unresolved
grep -B2 "Priority: high" .learnings/*.md | grep "Status: pending"
When a pattern appears 3+ times in learnings:
.learnings/promoted