Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Mine ALL past Claude conversations to build a living 'User Manual About You'. Extract writing style, business context, goals, preferences, and patterns. Make all other skills smarter with context.
.claude/skills/onewave-ai-conversation-archaeologist/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | 180% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 417% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 380% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 297% | 0% |
| case-10 | ✓→✗ | ▼ Worse | 484% | 0% |
Mine ALL past Claude conversations to build a living 'User Manual About You'. Extract writing style, business context, goals, preferences, and patterns. Make all other skills smarter with context.
You are a master conversation analyst and profile builder. Use conversation_search and recent_chats tools to mine hundreds of past conversations. Extract patterns in: writing style, business context, recurring problems, stated goals, preferences, pet peeves, domain expertise, relationship dynamics, and decision-making patterns. Create a comprehensive, living profile that other skills can reference for personalized outputs. Update this profile automatically as new conversations occur.
markdown# Conversation Archaeologist Output **Generated**: {timestamp} --- ## Results [Your formatted output here] --- ## Recommendations [Actionable next steps]
Trigger Phrases:
Example Request: > "Sample user request here]"
Response Approach:
Remember: Focus on delivering value quickly and clearly!
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-13 | fail→fail | 2,665 | 9,455 | +255% | 1 | 1 | 0% | 413 | 1,737 | +321% | 0 | 0 | — |
case-01 | fail→fail | 5,994 | 24,044 | +301% | 1 | 1 | 0% | 825 | 2,433 | +195% | 0 | 0 | — |
case-02 | fail→fail | 10,728 | 14,901 | +39% | 1 | 1 | 0% | 1,468 | 2,357 | +61% | 0 | 0 | — |
case-03 | fail→fail | 7,334 | 12,306 | +68% | 1 | 1 | 0% | 1,137 | 2,160 | +90% | 0 | 0 | — |
case-04 | pass→pass | 1,545 | 5,432 | +252% | 1 | 1 | 0% | 252 | 1,171 | +365% | 0 | 0 | — |
case-05 | pass→pass | 3,259 | 3,101 | -5% | 1 | 1 | 0% | 391 | 865 | +121% | 0 | 0 | — |
case-06 | pass→pass | 2,400 | 5,251 | +119% | 1 | 1 | 0% | 279 | 998 | +258% | 0 | 0 | — |
case-07 | fail→fail | 16,821 | 13,761 | -18% | 1 | 1 | 0% | 2,370 | 2,590 | +9% | 0 | 0 | — |
case-08 | fail→fail | 3,482 | 12,622 | +262% | 1 | 1 | 0% | 427 | 2,184 | +411% | 0 | 0 | — |
case-09 | fail→pass | 4,145 | 10,401 | +151% | 1 | 1 | 0% | 603 | 1,689 | +180% | 0 | 0 | — |
case-10 | pass→fail | 2,932 | 12,684 | +333% | 1 | 1 | 0% | 383 | 2,238 | +484% | 0 | 0 | — |
case-11 | fail→fail | 4,938 | 18,530 | +275% | 1 | 1 | 0% | 672 | 2,728 | +306% | 0 | 0 | — |
case-12 | fail→fail | 14,023 | 15,066 | +7% | 1 | 1 | 0% | 1,977 | 2,569 | +30% | 0 | 0 | — |
case-14 | fail→fail | 4,170 | 12,955 | +211% | 1 | 1 | 0% | 650 | 2,293 | +253% | 0 | 0 | — |
case-15 | fail→pass | 3,174 | 13,955 | +340% | 1 | 1 | 0% | 407 | 2,106 | +417% | 0 | 0 | — |
case-16 | fail→fail | 4,531 | 12,985 | +187% | 1 | 1 | 0% | 647 | 2,127 | +229% | 0 | 0 | — |
case-17 | fail→pass | 3,423 | 13,382 | +291% | 1 | 1 | 0% | 461 | 2,214 | +380% | 0 | 0 | — |
case-18 | fail→fail | 3,446 | 16,249 | +372% | 1 | 1 | 0% | 480 | 2,705 | +464% | 0 | 0 | — |
case-19 | fail→fail | 3,672 | 7,284 | +98% | 1 | 1 | 0% | 498 | 730 | +47% | 0 | 0 | — |
case-20 | fail→fail | 14,204 | 12,662 | -11% | 1 | 1 | 0% | 2,343 | 2,276 | -3% | 0 | 0 | — |
case-21 | fail→fail | 6,192 | 12,818 | +107% | 1 | 1 | 0% | 926 | 2,310 | +149% | 0 | 0 | — |
case-22 | fail→pass | 4,590 | 15,705 | +242% | 1 | 1 | 0% | 621 | 2,464 | +297% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +14 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.