search-conversations

Extraction

python3 ~/.claude/skills/search-conversations/scripts/extract_conversations.py --days 3

| Option | Effect | |--------|--------| | --days N | Days from last activity (not today) | | --from-today | Days from today instead | | --all-projects | Cross-project (implies --from-today) | | --project /path | Filter to specific project | | --compact | No metadata, more conversation | | --min-exchanges N | Skip sessions with < N exchanges | | --ids abc,def | Fetch specific conversations | | --paths /path.jsonl | Direct file paths |

Output: Default shows files, tools, errors + conversation. --compact omits metadata.

Time modes: --days N counts from last activity (useful when returning to old projects). --all-projects or --from-today counts from today (calendar-based).

Workflow

1. Identify the lens from user intent (see Routing table below)

2. Extract recent context using the script with lens-appropriate parameters:

   python3 ~/.claude/skills/search-conversations/scripts/extract_conversations.py --days N [flags]
   

   Use the Parameters table to select --days, flags, and any supplementary data to gather.

3. Apply lens questions to analyze the extracted conversations

4. Deepen the search using what you learned from the initial extraction:

   • Extract additional timeframes with the script (--days 30, --all-projects)
   • Search for specific keywords, project names, or patterns that surfaced:

     qmd search "keyword from context" -c conversations -n 15 --files
     

   • Extract those specific paths: python3 ... --paths /found/conv.jsonl

   This step surfaces older context, related discussions, or cross-project patterns that complement the initial extraction.

Keep index updated: Run qmd update periodically to index new sessions.

---

Lenses

Routing

| User Says | Lens | |-----------|------| | "where were we", "recap" | restore-context | | "what I learned", "reflect" | extract-learnings | | "gaps", "struggling" | find-gaps | | "mentor", "review process" | review-process | | "retro", "project review" | run-retro | | "decisions", "CLAUDE.md" | extract-decisions | | "bad habits", "antipatterns" | find-antipatterns |

Parameters

| Lens | Days | Flags | Also Gather | |------|------|-------|-------------| | restore-context | 3 | — | git status, git log -10 | | extract-learnings | 14 | --all-projects --compact | — | | find-gaps | 30 | --all-projects --compact | — | | review-process | 14 | --all-projects --compact | recent git log | | run-retro | 30 | --project /path | full git history | | extract-decisions | 90 | --project /path | — | | find-antipatterns | 30 | --all-projects --compact | — |

--min-exchanges 2 or 3 filters out short sessions and reduces noise.

Core Questions

| Lens | Ask | |------|-----| | restore-context | What's unfinished? What were the next steps? | | extract-learnings | Where did understanding shift? What mistakes became lessons? | | find-gaps | What topics recur? Where is guidance needed repeatedly? | | review-process | Is there planning before coding? Is debugging systematic? | | run-retro | How did the solution evolve? What worked? What was painful? | | extract-decisions | What trade-offs were discussed? What was rejected and why? | | find-antipatterns | What mistakes repeat? What confusions persist? |

Follow-ups: find-gaps → suggest learn-anything. extract-decisions → suggest /updateclaudemd.

Supplementary Search Patterns

When recent extraction doesn't surface enough, use these qmd queries to find specific sessions:

| Lens | Query | |------|-------| | extract-learnings | qmd search "learned realized understand clicked" -c conversations -n 15 --files | | find-gaps | qmd search "confused struggling help with don't understand" -c conversations -n 15 --files | | extract-decisions | qmd search "decided chose instead of trade-off because" -c conversations -n 15 --files | | find-antipatterns | qmd search "again same mistake repeated forgot" -c conversations -n 15 --files |

---

Synthesis

Principles

1. Prioritize significance — 3-5 key findings, not exhaustive lists
2. Be specific — file paths, dates, project names
3. Make it actionable — every finding suggests a response
4. Show evidence — quotes or references
5. Keep it scannable — clear structure, no walls of text

Structure

## [Analysis Type]: [Scope]

### Summary
[2-3 sentences]

### Findings
[Organized by whatever fits: categories, timeline, severity]

### Patterns
[Cross-cutting observations]

### Recommendations
[Actionable next steps]

Length

Default: 300-500 words. Expand only when data warrants it.