mattgierhart/PRD-driven-context-engineering / .claude/skills/ghm-harvest
ghm-harvest
VerifiedExtracts durable insights from temp/ files to SoT during EPIC Phase E. Triggers at EPIC completion or explicit `/ghm-harvest` invocation. Outputs new SoT entries and archive manifest.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 4.5 KB |
| assets/harvest-log-template.md | markdown | 4.2 KB |
| references/examples.md | markdown | 5.2 KB |
| references/harvest-patterns.md | markdown | 7.6 KB |
references/examples.md
View on GitHub# Harvest Examples
**Purpose**: Good and bad patterns for harvesting temp/ insights to SoT.
---
## Good Example: Comprehensive Harvest
### Temp Files (Before)
```
temp/
epic-015-notes.md (8KB)
user-research-findings.md (12KB)
api-schema-draft.md (4KB)
pricing-discussion.md (6KB)
debug-session-2026-01-10.log (15KB)
```
### Harvest Decisions
**epic-015-notes.md → Archive**
- Reason: Timeline, blockers, daily progress (useful for retro)
- Action: Move to `archive/2026-01/epic-015-notes.md`
**user-research-findings.md → Extract & Delete**
- Extract: 3 pain points to CFD-090, CFD-091, CFD-092
- Extract: 2 customer quotes (used in CFD entries)
- Delete: After extraction (insights preserved)
**api-schema-draft.md → Extract & Delete**
- Extract: Final schema to API-046 (User Preferences API)
- Delete: Draft superseded by final spec
**pricing-discussion.md → Extract & Archive**
- Extract: Final decision to BR-097 (Pricing Model: $50/mo)
- Archive: Team discussion (useful context for why we chose $50)
**debug-session-2026-01-10.log → Delete**
- Reason: Debugging logs (issue resolved, no future value)
### SoT Updates
**Created**:
- CFD-090: Onboarding confusion (Tier 1 evidence)
- CFD-091: Export feature request (Tier 2 evidence)
- CFD-092: Mobile app demand (Tier 1 evidence)
- BR-097: Pricing Model
**Updated**:
- API-046: Added final schema (from draft)
### Post-Harvest
**temp/**:
- Before: 5 files, 45KB
- After: 0 files, 0KB ✅
**SoT**:
- New IDs: 4
- Updated IDs: 1
**Archive**:
- epic-015-notes.md
- pricing-discussion.md
---
## Bad Example 1: No Extraction (Lost Insights)
### Temp Files (Before)
```
temp/
user-interviews.md (20KB - contains 5 interviews with pain points)
competitive-analysis.md (10KB - detailed competitor breakdown)
```
### Harvest (Bad)
**Action**: Deleted both files without review
**Result**:
- ❌ Lost 5 user pain points (no CFD entries created)
- ❌ Lost competitive insights (no CFD entries)
- ❌ Can't recover (files deleted)
### Fix
**Should Have**:
1. Reviewed user-interviews.md
2. Extracted each pain point to CFD-XXX
3. THEN deleted temp file
---
## Bad Example 2: Contaminating SoT
### Temp File
```markdown
# temp/pricing-decision.md
Team meeting 2026-01-10:
- Sarah suggested $50/mo
- John said $30/mo
- Discussed for 2 hours
- Looked at competitors: Asana ($30), Monday ($40), ClickUp ($35)
- Debated value vs cost
- Sarah won argument
- Final: $50/mo
```
### Bad Extraction
```markdown
# BR-097: Pricing Model
Team met on 2026-01-10. Sarah suggested $50/mo but John said $30/mo.
We discussed for 2 hours. We looked at competitors: Asana ($30), Monday ($40), ClickUp ($35).
We debated value vs cost. Sarah won the argument.
Final decision: $50/mo.
```
**Why Bad**: SoT contaminated with process, debate, names (not reusable)
### Good Extraction
**SoT (BR-097)**:
```markdown
# BR-097: Pricing Model
**Decision**: $50/month (per user)
**Rationale**:
- Positioned above mid-market ($30-40 range)
- Justifies premium features (advanced reporting, integrations)
- Competitive with high-end ($50-60 range: Asana Business)
**Alternatives Considered**:
- $30/mo: Too low (signals lower quality)
- $70/mo: Too high (limited market)
```
**Archive (pricing-discussion.md)**:
```markdown
[Full team discussion, debate, timeline]
Archived for: Retrospective on how we make pricing decisions
```
---
## Good Example: Selective Archiving
### Temp Files
```
temp/
epic-020-timeline.md (progress log)
epic-020-retro-notes.md (lessons learned)
temp-calculation.txt (intermediate math)
```
### Decisions
**epic-020-timeline.md → Archive**
- Reason: Useful for post-mortem (blockers, how long tasks took)
- Action: `archive/2026-01/epic-020-timeline.md`
**epic-020-retro-notes.md → Extract & Delete**
- Extract: Lessons to team wiki or EPIC template updates
- Example: "In-app prompts drive 3x adoption vs email" → Update GTM playbook
- Delete: After extracting lessons
**temp-calculation.txt → Delete**
- Reason: Intermediate work (final number in SoT, don't need steps)
---
## Bad Example 3: Over-Archiving
### Bad Pattern
**Archive Everything "Just in Case"**:
```
archive/2026-01/
debug-log-1.txt
debug-log-2.txt
scratch-notes.md
random-thoughts.md
temp-file-1.md
temp-file-2.md
[50 more files]
```
**Why Bad**:
- Archive becomes junk drawer
- Hard to find useful content
- No discrimination (noise alongside signal)
### Good Pattern
**Be Selective**:
```
archive/2026-01/
epic-020-timeline.md (retro value)
pricing-discussion.md (decision context)
ab-test-results.md (future learning)
```
**Delete the Rest**: Debug logs, scratch notes, intermediate work
---
## Harvest Checklist (Good Pattern)
### Before Harvest
- [ ] List all temp/ files
- [ ] Scan each for extractable insights
- [ ] Categorize: Extract | Archive | Delete
### During Harvest
- [ ] Extract durable insights to SoT
- [ ] Create/update IDs with proper cross-references
- [ ] Move contextual files to archive (with manifest entry)
- [ ] Delete noise (with confirmation)
### After Harvest
- [ ] temp/ directory empty ✅
- [ ] SoT updated (all insights preserved) ✅
- [ ] Archive documented (manifest) ✅
- [ ] Harvest log complete ✅
---
*Reference: Use these examples when harvesting insights in `ghm-harvest` skill.*
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |