Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Triage open herdr GitHub issues into a concise decision-first Markdown table. Use when the user says "triage", asks to triage open issues, asks which issues need attention, or wants issue priority/recommendation lights for herdr.
.claude/skills/herdrdev-triage/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -27% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -50% | 0% |
Use this skill only inside the herdr repository.
When the user says triage, inspect open GitHub issues for herdrdev/herdr and return a concise Markdown table. Prefer GitHub MCP tools when available. If they are unavailable, use gh issue list / gh issue view only when authenticated access is already configured.
Use this table shape:
| Light | Recommendation | Issue | Age | Reactions | Why | |---|---|---|---:|---:|---| | 🔴 | fix now | #123 | 18d | 5 | user-visible regression | | 🟡 | queue | #124 | 42d | 2 | useful but not blocking | | 🔵 | defer | #125 | 7d | 0 | cosmetic polish |
Keep issue numbers as Markdown links. Use days since issue creation for Age. Use total reactions for Reactions; include a compact breakdown only when it changes interpretation, such as 7 (5 👍, 2 👀).
Classify with these lights:
fix now: reproducible bug, crash, data loss, blocked workflow, release risk, or high-confidence user-visible regression.queue: useful feature, important quality issue, repeated user signal, stale issue that still looks valid, or behavior worth scheduling.defer: cosmetic polish, low-signal idea, unclear report, docs-only nit, or issue that likely needs more evidence before implementation.Recommendations should be short imperative phrases: fix now, queue, defer, needs repro, close?, or needs owner decision.
Write one sentence before the table only if needed to state scope, such as how many open issues were inspected. After the table, add at most one short note for uncertainty or follow-up. Do not produce a long narrative unless the user asks for depth.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 8,704 | 4,394 | -50% | 1 | 1 | 0% | 1,733 | 734 | -58% | 0 | 0 | — |
case-02 | fail→fail | 12,419 | 6,777 | -45% | 1 | 1 | 0% | 2,513 | 937 | -63% | 0 | 0 | — |
case-03 | fail→fail | 13,266 | 26,337 | +99% | 1 | 1 | 0% | 2,334 | 5,833 | +150% | 0 | 0 | — |
case-04 | fail→fail | 11,191 | 4,950 | -56% | 1 | 1 | 0% | 1,829 | 780 | -57% | 0 | 0 | — |
case-05 | fail→fail | 8,970 | 7,285 | -19% | 1 | 1 | 0% | 1,566 | 939 | -40% | 0 | 0 | — |
case-06 | fail→pass | 5,686 | 2,046 | -64% | 1 | 1 | 0% | 945 | 855 | -10% | 0 | 0 | — |
case-07 | fail→pass | 4,976 | 2,685 | -46% | 1 | 1 | 0% | 838 | 911 | +9% | 0 | 0 | — |
case-08 | fail→pass | 6,055 | 3,194 | -47% | 1 | 1 | 0% | 1,052 | 1,093 | +4% | 0 | 0 | — |
case-09 | fail→pass | 8,074 | 2,553 | -68% | 1 | 1 | 0% | 1,320 | 960 | -27% | 0 | 0 | — |
case-10 | pass→fail | 7,039 | 5,902 | -16% | 1 | 1 | 0% | 1,291 | 831 | -36% | 0 | 0 | — |
case-11 | pass→pass | 6,666 | 1,625 | -76% | 1 | 1 | 0% | 1,205 | 700 | -42% | 0 | 0 | — |
case-12 | fail→pass | 8,509 | 1,857 | -78% | 1 | 1 | 0% | 1,543 | 769 | -50% | 0 | 0 | — |
case-13 | fail→fail | 8,366 | 5,372 | -36% | 1 | 1 | 0% | 1,513 | 797 | -47% | 0 | 0 | — |
case-14 | fail→fail | 7,143 | 7,221 | +1% | 1 | 1 | 0% | 1,273 | 1,076 | -15% | 0 | 0 | — |
case-15 | fail→pass | 6,974 | 2,937 | -58% | 1 | 1 | 0% | 1,255 | 1,121 | -11% | 0 | 0 | — |
case-16 | fail→pass | 7,157 | 2,971 | -58% | 1 | 1 | 0% | 1,286 | 1,053 | -18% | 0 | 0 | — |
case-17 | pass→pass | 9,989 | 3,027 | -70% | 1 | 1 | 0% | 2,055 | 1,005 | -51% | 0 | 0 | — |
case-18 | fail→pass | 13,522 | 3,367 | -75% | 1 | 1 | 0% | 2,151 | 911 | -58% | 0 | 0 | — |
case-19 | fail→pass | 8,101 | 2,269 | -72% | 1 | 1 | 0% | 1,328 | 781 | -41% | 0 | 0 | — |
case-20 | pass→fail | 8,331 | 7,669 | -8% | 1 | 1 | 0% | 768 | 1,786 | +133% | 0 | 0 | — |
case-21 | pass→pass | 9,020 | 8,731 | -3% | 1 | 1 | 0% | 1,777 | 2,256 | +27% | 0 | 0 | — |
case-22 | pass→fail | 11,383 | 3,875 | -66% | 1 | 1 | 0% | 1,896 | 929 | -51% | 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 16 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 +27 percentage points is the difference between those two pass rates over the 16 comparable cases. 4 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.