Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Persistent memory CLI for LLM agents. Store facts, recall past knowledge, link related memories, manage lifecycle.
.claude/skills/mnemon-dev-mnemon/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -6% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 195% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -55% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -24% | 0% |
mnemon remember "<fact>" --cat <cat> --imp <1-5> --entities "e1,e2" --source agentaction (added/updated/skipped), semantic_candidates, and causal_candidates.causal_candidates: link only when the memories are genuinely causally related.semantic_candidates: high similarity alone is not enough; skip unrelated keyword matches.mnemon link <id> <candidate> --type <causal|semantic> --weight <0-1> [--meta '<json>']mnemon recall "<query>" --limit 10bashmnemon remember "<fact>" --cat <cat> --imp <1-5> --entities "e1,e2" --source agent mnemon link <id1> <id2> --type <type> --weight <0-1> [--meta '<json>'] mnemon recall "<query>" --limit 10 mnemon search "<query>" --limit 10 mnemon import --dry-run <file> mnemon import <file> mnemon forget <id> mnemon related <id> --edge causal mnemon gc --threshold 0.4 mnemon gc --keep <id> mnemon status mnemon log mnemon store list mnemon store create <name> mnemon store set <name> mnemon store remove <name>
When the user asks to import old chats, notes, or exported context, create a memory_draft.json with schema_version: "1", insights entries containing content, category, importance, tags, entities, and optional created_at, plus optional edges using source_index, target_index, edge_type, weight, and reason. Run mnemon import --dry-run <file>, then run mnemon import <file> only after validation passes. After import, verify with mnemon status and a focused mnemon search or mnemon recall. Check the output errors field because imports can partially succeed.
preference · decision · insight · fact · contexttemporal · semantic · causal · entity| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 7,466 | 4,896 | -34% | 1 | 1 | 0% | 340 | 996 | +193% | 0 | 0 | — |
case-02 | fail→fail | 5,144 | 12,116 | +136% | 1 | 1 | 0% | 614 | 924 | +50% | 0 | 0 | — |
case-03 | fail→fail | 4,395 | 6,393 | +45% | 1 | 1 | 0% | 710 | 1,099 | +55% | 0 | 0 | — |
case-04 | pass→pass | 8,344 | 6,385 | -23% | 1 | 1 | 0% | 1,490 | 1,759 | +18% | 0 | 0 | — |
case-05 | pass→pass | 9,807 | 5,823 | -41% | 1 | 1 | 0% | 1,807 | 1,675 | -7% | 0 | 0 | — |
case-06 | pass→pass | 7,241 | 6,638 | -8% | 1 | 1 | 0% | 1,356 | 1,750 | +29% | 0 | 0 | — |
case-07 | pass→pass | 4,518 | 2,847 | -37% | 1 | 1 | 0% | 729 | 1,113 | +53% | 0 | 0 | — |
case-08 | fail→pass | 8,428 | 1,958 | -77% | 1 | 1 | 0% | 1,439 | 991 | -31% | 0 | 0 | — |
case-09 | fail→pass | 7,562 | 8,765 | +16% | 1 | 1 | 0% | 1,273 | 1,191 | -6% | 0 | 0 | — |
case-10 | fail→pass | 4,678 | 12,254 | +162% | 1 | 1 | 0% | 792 | 2,340 | +195% | 0 | 0 | — |
case-11 | fail→fail | 5,931 | 6,556 | +11% | 1 | 1 | 0% | 999 | 1,005 | +1% | 0 | 0 | — |
case-12 | fail→pass | 23,238 | 1,928 | -92% | 1 | 1 | 0% | 2,325 | 1,045 | -55% | 0 | 0 | — |
case-13 | fail→pass | 7,753 | 1,426 | -82% | 1 | 1 | 0% | 1,156 | 875 | -24% | 0 | 0 | — |
case-14 | fail→pass | 6,526 | 1,476 | -77% | 1 | 1 | 0% | 1,063 | 844 | -21% | 0 | 0 | — |
case-15 | fail→fail | 8,031 | 5,250 | -35% | 1 | 1 | 0% | 1,552 | 962 | -38% | 0 | 0 | — |
case-16 | fail→pass | 5,359 | 6,143 | +15% | 1 | 1 | 0% | 1,056 | 1,932 | +83% | 0 | 0 | — |
case-17 | fail→pass | 18,951 | 2,049 | -89% | 1 | 1 | 0% | 3,347 | 983 | -71% | 0 | 0 | — |
case-18 | fail→pass | 14,441 | 3,500 | -76% | 1 | 1 | 0% | 2,650 | 1,301 | -51% | 0 | 0 | — |
case-19 | fail→fail | 5,631 | 4,283 | -24% | 1 | 1 | 0% | 943 | 887 | -6% | 0 | 0 | — |
case-20 | fail→pass | 15,156 | 2,400 | -84% | 1 | 1 | 0% | 2,611 | 960 | -63% | 0 | 0 | — |
case-21 | fail→pass | 10,913 | 1,783 | -84% | 1 | 1 | 0% | 2,062 | 938 | -55% | 0 | 0 | — |
case-22 | fail→pass | 5,667 | 3,837 | -32% | 1 | 1 | 0% | 1,210 | 1,514 | +25% | 0 | 0 | — |
case-23 | fail→pass | 11,556 | 4,676 | -60% | 1 | 1 | 0% | 1,905 | 1,496 | -21% | 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. 23 cases were attempted, and 17 counted toward the lift figure. The other 6 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 +57 percentage points is the difference between those two pass rates over the 17 comparable cases.
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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/22/2026 | +52% |
Other measured skills in the registry, with their headline benchmark lift.