triage-perf-verdict
Skill1 starsChanged 3 months agoMITView on GitHub
Type: skill (1 model narration) · Used by: perf-run-one · Status: LIVE You will draft a Jira regression ticket for a red performance verdict. Follow these steps exactly. Inputs (passed by the workflow): - run_id — the run identifier - grafana_url — deep-link URL from O7 link-grafana-panel; empty string "" if O7 was skipped or OBSERVE!=1 Read both files: - reports/<run_id>/verdict.json - reports/<run_id>/contract.json From the verdict and contract, write exactly one sentence describing what regressed and by how much. Rules:…
In this repository42 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- k6-browser-testSkill
- k6-load-testSkill
- k6-playwright-converterSkill
- k6-smoke-testSkill
- k6-test-plannerSkill
- apply-ws2-conventionsSkill
- author-resilient-playwrightSkill
- build-status-viewsSkill
- choose-k6-execution-modeSkill
- compare-coreSkill
- compare-to-baselineSkill
- corroborate-2-sourcesSkill
- curate-baselinesSkill
- doc-consistency-checkSkill
- emit-otlp-and-prometheusSkill
- estimate-token-costSkill
- explore-product-structureSkill
- file-perf-regression-jiraSkill
- func-verdictSkill
- heal-playwright-suiteSkill
- k6-profile-benchmarkSkill
- k6-profile-loadSkill
- k6-profile-soakSkill
- k6-profile-spikeSkill
- k6-profile-stressSkill
- learn-via-k6-docsSkill
- link-grafana-panelSkill
- notify-responsible-teamSkill
- open-draft-prSkill
- parse-k6-json-summarySkill
- parse-playwright-summarySkill
- perf-run-logSkill
- run-k6-scriptSkill
- run-ledgerSkill
- run-playwright-suiteSkill
- scope-reviewSkill
- scrub-har-secretsSkill
- triage-func-verdictSkill
- verify-k6-scriptSkill
- verify-playwright-suiteSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 29
If multiple endpoints are red, name all of them in the same sentence.
- Very specific: line 34
All fields except [NARRATION] are deterministic — copy the values verbatim, no invention.
- Very specific: line 72
Write the assembled draft to reports/<run_id>/jira-draft.md.
- Distinctive: line 86
Read and Write only — one model narration call, no MCP.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
triage-perf-verdict
Draft Jira ticket text for a red verdict — 1 AI narration line + deterministic template.
Type: skill (1 model narration) · Used by: perf-run-one · Status: LIVE
Prompt
You will draft a Jira regression ticket for a red performance verdict. Follow these steps exactly.
1 — Read inputs
Inputs (passed by the workflow):
run_id— the run identifiergrafana_url— deep-link URL from O7link-grafana-panel; empty string""if O7 was skipped or OBSERVE!=1
Read both files:
reports/<run_id>/verdict.jsonreports/<run_id>/contract.json
2 — Write ONE narration line (model call)
From the verdict and contract, write exactly one sentence describing what regressed and by how much. Rules:
- Neutral tone. No blame, no guessing at root cause.
- Name the endpoint, the p95, the threshold, and the delta.
- Example: "GET /api/config p95 climbed to 1250ms (+350ms, +38.9% over the 900ms red threshold) on demo-web STG as of 2026-06-09."
- If multiple endpoints are red, name all of them in the same sentence.
3 — Assemble the draft ticket
Fill the template below with values from the verdict/contract + the narration line.
All fields except [NARRATION] are deterministic — copy the values verbatim, no invention.
h2. Performance Regression — Auto-detected
Team: <team>
Env: <env>
Run ID: <run_id>
Verdict: red
Date: <started_at from contract, date part only>
Summary: [NARRATION]
h3. Failing endpoints
| Endpoint | p95 ms (Result) | Red threshold ms | Delta |
| <name> | <p95_ms> | <p95_red_ms> | +<ms> (+<pct>%) |
h3. Checks
- Total iterations: <contract.iterations>
- Check pass rate: <contract.checks.pass_rate × 100>%
- Error rate: <endpoint error_rate>
h3. Links
- Grafana: <grafana_url if non-empty, else "[unavailable]">
- Run ledger: state/run-ledger.jsonl (filter run_id=<run_id>)
h3. Action required
Human review required before this ticket is filed. Route to reviewer subagent next.Template is Jira wiki markup (h2./h3., pipe-delimited tables) — NOT markdown. This is the format that file-perf-regression-jira copies verbatim into the Jira description field.
4 — Write output
Write the assembled draft to reports/<run_id>/jira-draft.md.
Return:
{
"draft_path": "reports/<run_id>/jira-draft.md",
"narration": "<the one-sentence narration>"
}Tools
Read and Write only — one model narration call, no MCP.
Data
- Inputs:
run_id(string),grafana_url(string — from O7; empty string if unavailable) - Reads:
reports/<run_id>/verdict.json,reports/<run_id>/contract.json - Writes:
reports/<run_id>/jira-draft.md
Hard rules
- Exactly one model narration line — the rest of the ticket is templated from the data.
- Draft only —
file-perf-regression-jirais the separate human-gated filing step. - Only runs on red verdicts — green and no-baseline never reach this skill.
- p95 everywhere — never quote p50 or p99 in the narration.
Frozen skill registry: AGENTS.md in this repo
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

