Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Build and query AI-powered knowledge bases from claude-mem observations. Use when users want to create focused "brains" from their observation history, ask questions about past work patterns, or compile expertise on specific topics.
.claude/skills/thedotmack-knowledge-agent/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -34% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -58% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 2% | 0% |
Build and query AI-powered knowledge bases from claude-mem observations.
Knowledge agents are filtered corpora of observations compiled into a conversational AI session. Build a corpus from your observation history, prime it (loads the knowledge into an AI session), then ask it questions conversationally.
Think of them as custom "brains": "everything about hooks", "all decisions from the last month", "all bugfixes for the worker service".
textbuild_corpus name="hooks-expertise" description="Everything about the hooks lifecycle" project="claude-mem" concepts="hooks" limit=500
Filter options:
project — filter by project nametypes — comma-separated: decision, bugfix, feature, refactor, discovery, changeconcepts — comma-separated concept tagsfiles — comma-separated file paths (prefix match)query — semantic search querydateStart / dateEnd — ISO date rangelimit — max observations (default 500)textprime_corpus name="hooks-expertise"
This creates an AI session loaded with all the corpus knowledge. Takes a moment for large corpora.
textquery_corpus name="hooks-expertise" question="What are the 5 lifecycle hooks and when does each fire?"
The knowledge agent answers from its corpus. Follow-up questions maintain context.
textlist_corpora
Shows all corpora with stats and priming status.
textrebuild_corpus name="hooks-expertise"
After rebuilding, reprime to load the updated knowledge:
textreprime_corpus name="hooks-expertise"
Clears prior Q&A context and reloads the corpus into a new session.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 17,321 | 11,058 | -36% | 1 | 1 | 0% | 2,322 | 2,634 | +13% | 0 | 0 | — |
case-02 | fail→fail | 6,901 | 3,115 | -55% | 1 | 1 | 0% | 1,097 | 818 | -25% | 0 | 0 | — |
case-03 | fail→fail | 8,811 | 12,509 | +42% | 1 | 1 | 0% | 1,513 | 1,269 | -16% | 0 | 0 | — |
case-04 | fail→pass | 9,842 | 8,220 | -16% | 1 | 1 | 0% | 1,810 | 1,194 | -34% | 0 | 0 | — |
case-05 | fail→pass | 18,962 | 9,325 | -51% | 1 | 1 | 0% | 3,044 | 1,666 | -45% | 0 | 0 | — |
case-06 | fail→fail | 7,320 | 3,508 | -52% | 1 | 1 | 0% | 1,459 | 844 | -42% | 0 | 0 | — |
case-07 | fail→fail | 13,168 | 2,434 | -82% | 1 | 1 | 0% | 3,108 | 761 | -76% | 0 | 0 | — |
case-08 | fail→pass | 14,591 | 3,356 | -77% | 1 | 1 | 0% | 2,737 | 1,141 | -58% | 0 | 0 | — |
case-09 | fail→pass | 4,498 | 1,879 | -58% | 1 | 1 | 0% | 841 | 862 | +2% | 0 | 0 | — |
case-10 | fail→pass | 6,827 | 4,684 | -31% | 1 | 1 | 0% | 1,210 | 1,429 | +18% | 0 | 0 | — |
case-11 | fail→pass | 3,619 | 2,752 | -24% | 1 | 1 | 0% | 690 | 1,020 | +48% | 0 | 0 | — |
case-12 | fail→pass | 2,315 | 1,332 | -42% | 1 | 1 | 0% | 389 | 763 | +96% | 0 | 0 | — |
case-13 | fail→pass | 10,356 | 2,171 | -79% | 1 | 1 | 0% | 1,843 | 641 | -65% | 0 | 0 | — |
case-18 | fail→pass | 7,493 | 4,448 | -41% | 1 | 1 | 0% | 1,821 | 883 | -52% | 0 | 0 | — |
case-14 | fail→pass | 5,770 | 5,148 | -11% | 1 | 1 | 0% | 1,213 | 1,453 | +20% | 0 | 0 | — |
case-15 | fail→pass | 8,054 | 1,810 | -78% | 1 | 1 | 0% | 1,652 | 764 | -54% | 0 | 0 | — |
case-16 | fail→fail | 9,763 | 6,037 | -38% | 1 | 1 | 0% | 1,945 | 863 | -56% | 0 | 0 | — |
case-17 | fail→pass | 9,923 | 6,571 | -34% | 1 | 1 | 0% | 1,961 | 1,045 | -47% | 0 | 0 | — |
case-19 | fail→pass | 8,248 | 4,484 | -46% | 1 | 1 | 0% | 1,562 | 1,295 | -17% | 0 | 0 | — |
case-20 | pass→pass | 3,644 | 4,954 | +36% | 1 | 1 | 0% | 713 | 1,469 | +106% | 0 | 0 | — |
case-21 | pass→pass | 7,244 | 6,026 | -17% | 1 | 1 | 0% | 1,343 | 1,606 | +20% | 0 | 0 | — |
case-22 | pass→pass | 2,694 | 11,280 | +319% | 1 | 1 | 0% | 483 | 1,596 | +230% | 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 20 counted toward the lift figure. The other 2 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 +64 percentage points is the difference between those two pass rates over the 20 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.