Triage complet RTK : exécute issue-triage + pr-triage en parallèle, puis croise les données pour détecter doubles couvertures, trous sécurité, P0 sans PR, et conflits internes. Sauvegarde dans claudedocs/RTK-YYYY-MM-DD.md. Args: "en"/"fr" pour la langue (défaut: fr), "save" pour forcer la sauvegarde.
SkillScore77Maintained
Automatic
◆DevOps & SRE pack
Last verified 8/7/2026
What it changes
Passed 16 of 21 test cases with the skill, and 11 of 21 without it.
Without skill50%
With skill77%
Checks it passed that the model alone missed
Saves the output file to claudedocs/RTK-YYYY-MM-DD.md by default
Notifies the user when open items exceed the 200 fetching threshold
Suppresses report saving only when explicitly instructed with no save
and 5 more in the full report
6 fixed11 passed either way5 still failing0 broke
Pass-rate lift
+27 pp
Tokens
+46%
Agent turns
0%
Response time
-55%
Model gemini-3.6-flashMeasured 8/7/2026v117 of 22 cases passed with skill · 21 comparable
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 6 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 6
Prompt
When completing a full cross-triage run on open items, where is the generated analysis report stored by default, and under what condition should file saving be skipped?
What was checked, without and with the skill
Checked
Without skill
With skill
Saves the output file to claudedocs/RTK-YYYY-MM-DD.md by default
Missed
Passed
Suppresses report saving only when explicitly instructed with no save
Missed
Passed
case-16 · gemini-3.6-flash · 8/7/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md228 linessingle file
/rtk-triage
Orchestrateur de triage RTK. Fusionne issue-triage + pr-triage et produit une analyse croisée.
Quand utiliser
Hebdomadaire ou avant chaque sprint
Quand le backlog PR/issues grossit rapidement
Pour identifier les doublons avant de reviewer
Workflow en 4 phases
Phase 0 — Préconditions
git rev-parse --is-inside-work-tree
gh auth status
Vérifier que la date actuelle est connue (utiliser date +%Y-%m-%d).
Phase 1 — Data gathering (parallèle)
Lancer les deux collectes simultanément :
Issues :
gh repo view --json nameWithOwner -q .nameWithOwner
gh issue list --state open --limit 150 \
--json number,title,author,createdAt,updatedAt,labels,assignees,body
gh issue list --state closed --limit 20 \
--json number,title,labels,closedAt