Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Quick summary of the last session — commands run, files changed, and what to do next.
.claude/skills/sharpdeveye-recap/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-18 | ✗→✓ | ▲ Improved | -16% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -47% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 356% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 489% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 104% | 0% |
Invoke /agent-workflow — it contains workflow principles, anti-patterns, and the Context Gathering Protocol. Follow the protocol before proceeding — if no workflow context exists yet, you MUST run /teach-maestro first.
Get back up to speed fast. Recap reads the most recent session file and recent decisions to tell you where you left off and what to do next.
.maestro/sessions/ — read the most recent session file (by filename date).maestro/decisions.jsonl — read the last 5 decisionsIf no session files exist, read the last 10 decisions and summarize those instead.
If no data exists at all, respond: "No session history found. Run `/capture` at the end of your next session to start tracking."
text┌─────────────────────────────────────┐ │ LAST SESSION RECAP │ ├─────────────────────────────────────┤ │ Date: YYYY-MM-DD │ │ Commands: /diagnose → /fortify (2) │ │ Files: 3 changed, 1 created │ │ Outcome: Completed │ ├─────────────────────────────────────┤ │ PICK UP HERE: │ │ 1. Run /guard — rate limiting │ │ 2. Run /evaluate — adversarial tests│ └─────────────────────────────────────┘
After recapping, run the first command listed in "Pick Up Here" to continue where you left off.
NEVER:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→fail | 3,976 | 5,495 | +38% | 1 | 1 | 0% | 580 | 736 | +27% | 0 | 0 | — |
case-01 | fail→fail | 3,689 | 35,177 | +854% | 1 | 1 | 0% | 512 | 765 | +49% | 0 | 0 | — |
case-03 | fail→fail | 3,554 | 5,930 | +67% | 1 | 1 | 0% | 456 | 721 | +58% | 0 | 0 | — |
case-04 | pass→fail | 3,019 | 6,937 | +130% | 1 | 1 | 0% | 401 | 819 | +104% | 0 | 0 | — |
case-05 | fail→fail | 6,074 | 18,101 | +198% | 1 | 1 | 0% | 842 | 3,743 | +345% | 0 | 0 | — |
case-06 | fail→fail | 4,426 | 7,423 | +68% | 1 | 1 | 0% | 754 | 1,739 | +131% | 0 | 0 | — |
case-18 | fail→pass | 11,307 | 15,721 | +39% | 1 | 1 | 0% | 1,840 | 1,537 | -16% | 0 | 0 | — |
case-07 | fail→pass | 9,345 | 2,488 | -73% | 1 | 1 | 0% | 1,531 | 813 | -47% | 0 | 0 | — |
case-08 | fail→fail | 9,355 | 10,329 | +10% | 1 | 1 | 0% | 1,362 | 908 | -33% | 0 | 0 | — |
case-09 | fail→pass | 3,409 | 8,558 | +151% | 1 | 1 | 0% | 460 | 2,099 | +356% | 0 | 0 | — |
case-10 | fail→fail | 10,645 | 5,771 | -46% | 1 | 1 | 0% | 1,777 | 665 | -63% | 0 | 0 | — |
case-11 | pass→pass | 7,776 | 3,118 | -60% | 1 | 1 | 0% | 1,168 | 939 | -20% | 0 | 0 | — |
case-12 | fail→fail | 2,624 | 7,823 | +198% | 1 | 1 | 0% | 372 | 972 | +161% | 0 | 0 | — |
case-13 | fail→pass | 7,172 | 12,772 | +78% | 1 | 1 | 0% | 366 | 2,155 | +489% | 0 | 0 | — |
case-14 | pass→fail | 5,051 | 9,114 | +80% | 1 | 1 | 0% | 530 | 737 | +39% | 0 | 0 | — |
case-15 | fail→fail | 8,931 | 7,755 | -13% | 1 | 1 | 0% | 1,300 | 932 | -28% | 0 | 0 | — |
case-16 | fail→fail | 10,254 | 7,320 | -29% | 1 | 1 | 0% | 1,577 | 752 | -52% | 0 | 0 | — |
case-17 | pass→pass | 7,356 | 10,239 | +39% | 1 | 1 | 0% | 1,287 | 1,705 | +32% | 0 | 0 | — |
case-19 | fail→fail | 3,210 | 5,152 | +60% | 1 | 1 | 0% | 389 | 710 | +83% | 0 | 0 | — |
case-20 | pass→fail | 10,049 | 25,451 | +153% | 1 | 1 | 0% | 1,850 | 5,011 | +171% | 0 | 0 | — |
case-21 | pass→pass | 5,302 | 14,880 | +181% | 1 | 1 | 0% | 849 | 2,360 | +178% | 0 | 0 | — |
case-22 | pass→pass | 18,496 | 14,623 | -21% | 1 | 1 | 0% | 3,075 | 2,616 | -15% | 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 9 counted toward the lift figure. The other 13 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 +5 percentage points is the difference between those two pass rates over the 9 comparable cases. 3 cases got worse with the skill loaded, and they are 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.