Check Skills workflow (trpc/trpc)
The Check Skills workflow from trpc/trpc, explained and optimized by Latchkey.
CI health: C - fair
Point runs-on at Latchkey and get caching, job timeouts, self-healing for flaky steps, and up to 58% lower cost, applied automatically.
What it does
This is the Check Skills workflow from the trpc/trpc repository, a real project running GitHub Actions. It is shown here with attribution under its MIT license.
Below, Latchkey shows a faster, safer version produced by its optimization engine.
The workflow
# check-skills.yml - Drop this into your library repo's .github/workflows/
#
# Checks for stale intent skills after a release and opens a review PR
# if any skills need attention. The PR body includes a prompt you can
# paste into Claude Code, Cursor, or any coding agent to update them.
#
# Triggers: new release published, or manual workflow_dispatch.
#
# Template variables (replaced by `intent setup`):
# @trpc/trpc - e.g. @tanstack/query or my-workspace workspace
name: Check Skills
on:
release:
types: [published]
workflow_dispatch: {}
permissions:
contents: write
pull-requests: write
jobs:
check:
name: Check for stale skills
runs-on: ubuntu-latest
steps:
- name: Checkout
uses: actions/checkout@v6
with:
fetch-depth: 0
- name: Setup Node
uses: actions/setup-node@v6
with:
node-version: 20
- name: Install intent
run: npm install -g @tanstack/intent
- name: Check staleness
id: stale
run: |
OUTPUT=$(intent stale --json 2>&1) || true
echo "$OUTPUT"
# Check if any skills need review
NEEDS_REVIEW=$(echo "$OUTPUT" | node -e "
const input = require('fs').readFileSync('/dev/stdin','utf8');
try {
const reports = JSON.parse(input);
const stale = reports.flatMap(r =>
r.skills.filter(s => s.needsReview).map(s => ({ library: r.library, skill: s.name, reasons: s.reasons }))
);
if (stale.length > 0) {
console.log(JSON.stringify(stale));
}
} catch {}
")
if [ -z "$NEEDS_REVIEW" ]; then
echo "has_stale=false" >> "$GITHUB_OUTPUT"
else
echo "has_stale=true" >> "$GITHUB_OUTPUT"
# Escape for multiline GH output
EOF=$(dd if=/dev/urandom bs=15 count=1 status=none | base64)
echo "stale_json<<$EOF" >> "$GITHUB_OUTPUT"
echo "$NEEDS_REVIEW" >> "$GITHUB_OUTPUT"
echo "$EOF" >> "$GITHUB_OUTPUT"
fi
- name: Build summary
if: steps.stale.outputs.has_stale == 'true'
id: summary
run: |
node -e "
const stale = JSON.parse(process.env.STALE_JSON);
const lines = stale.map(s =>
'- **' + s.skill + '** (' + s.library + '): ' + s.reasons.join(', ')
);
const summary = lines.join('\n');
const prompt = [
'Review and update the following stale intent skills for @trpc/trpc:',
'',
...stale.map(s => '- ' + s.skill + ': ' + s.reasons.join(', ')),
'',
'For each stale skill:',
'1. Read the current SKILL.md file',
'2. Check what changed in the library since the skill was last updated',
'3. Update the skill content to reflect current APIs and behavior',
'4. Run \`npx @tanstack/intent validate\` to verify the updated skill',
].join('\n');
// Write outputs
const fs = require('fs');
const env = fs.readFileSync(process.env.GITHUB_OUTPUT, 'utf8');
const eof = require('crypto').randomBytes(15).toString('base64');
fs.appendFileSync(process.env.GITHUB_OUTPUT,
'summary<<' + eof + '\n' + summary + '\n' + eof + '\n' +
'prompt<<' + eof + '\n' + prompt + '\n' + eof + '\n'
);
"
env:
STALE_JSON: ${{ steps.stale.outputs.stale_json }}
- name: Open review PR
if: steps.stale.outputs.has_stale == 'true'
env:
GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: |
VERSION="${{ github.event.release.tag_name || 'manual' }}"
BRANCH="skills/review-${VERSION}"
git config user.name "github-actions[bot]"
git config user.email "41898282+github-actions[bot]@users.noreply.github.com"
git checkout -b "$BRANCH"
git commit --allow-empty -m "chore: review stale skills for ${VERSION}"
git push origin "$BRANCH"
gh pr create \
--title "Review stale skills (${VERSION})" \
--body "$(cat <<'PREOF'
## Stale Skills Detected
The following skills may need updates after the latest release:
${{ steps.summary.outputs.summary }}
---
### Update Prompt
Paste this into your coding agent (Claude Code, Cursor, etc.):
~~~
${{ steps.summary.outputs.prompt }}
~~~
PREOF
)" \
--head "$BRANCH" \
--base main
The same workflow, on Latchkey
Estimated ~20% faster on cache hits, plus fewer wasted runs and a safer supply chain. Added and changed lines are highlighted.
# check-skills.yml - Drop this into your library repo's .github/workflows/ # # Checks for stale intent skills after a release and opens a review PR # if any skills need attention. The PR body includes a prompt you can # paste into Claude Code, Cursor, or any coding agent to update them. # # Triggers: new release published, or manual workflow_dispatch. # # Template variables (replaced by `intent setup`): # @trpc/trpc - e.g. @tanstack/query or my-workspace workspace name: Check Skills on: release: types: [published] workflow_dispatch: {} permissions: contents: write pull-requests: write jobs: check: timeout-minutes: 30 name: Check for stale skills runs-on: latchkey-small steps: - name: Checkout uses: actions/checkout@v6 with: fetch-depth: 0 - name: Setup Node uses: actions/setup-node@v6 with: cache: 'npm' node-version: 20 - name: Install intent run: npm install -g @tanstack/intent - name: Check staleness id: stale run: | OUTPUT=$(intent stale --json 2>&1) || true echo "$OUTPUT" # Check if any skills need review NEEDS_REVIEW=$(echo "$OUTPUT" | node -e " const input = require('fs').readFileSync('/dev/stdin','utf8'); try { const reports = JSON.parse(input); const stale = reports.flatMap(r => r.skills.filter(s => s.needsReview).map(s => ({ library: r.library, skill: s.name, reasons: s.reasons })) ); if (stale.length > 0) { console.log(JSON.stringify(stale)); } } catch {} ") if [ -z "$NEEDS_REVIEW" ]; then echo "has_stale=false" >> "$GITHUB_OUTPUT" else echo "has_stale=true" >> "$GITHUB_OUTPUT" # Escape for multiline GH output EOF=$(dd if=/dev/urandom bs=15 count=1 status=none | base64) echo "stale_json<<$EOF" >> "$GITHUB_OUTPUT" echo "$NEEDS_REVIEW" >> "$GITHUB_OUTPUT" echo "$EOF" >> "$GITHUB_OUTPUT" fi - name: Build summary if: steps.stale.outputs.has_stale == 'true' id: summary run: | node -e " const stale = JSON.parse(process.env.STALE_JSON); const lines = stale.map(s => '- **' + s.skill + '** (' + s.library + '): ' + s.reasons.join(', ') ); const summary = lines.join('\n'); const prompt = [ 'Review and update the following stale intent skills for @trpc/trpc:', '', ...stale.map(s => '- ' + s.skill + ': ' + s.reasons.join(', ')), '', 'For each stale skill:', '1. Read the current SKILL.md file', '2. Check what changed in the library since the skill was last updated', '3. Update the skill content to reflect current APIs and behavior', '4. Run \`npx @tanstack/intent validate\` to verify the updated skill', ].join('\n'); // Write outputs const fs = require('fs'); const env = fs.readFileSync(process.env.GITHUB_OUTPUT, 'utf8'); const eof = require('crypto').randomBytes(15).toString('base64'); fs.appendFileSync(process.env.GITHUB_OUTPUT, 'summary<<' + eof + '\n' + summary + '\n' + eof + '\n' + 'prompt<<' + eof + '\n' + prompt + '\n' + eof + '\n' ); " env: STALE_JSON: ${{ steps.stale.outputs.stale_json }} - name: Open review PR if: steps.stale.outputs.has_stale == 'true' env: GH_TOKEN: ${{ secrets.GITHUB_TOKEN }} run: | VERSION="${{ github.event.release.tag_name || 'manual' }}" BRANCH="skills/review-${VERSION}" git config user.name "github-actions[bot]" git config user.email "41898282+github-actions[bot]@users.noreply.github.com" git checkout -b "$BRANCH" git commit --allow-empty -m "chore: review stale skills for ${VERSION}" git push origin "$BRANCH" gh pr create \ --title "Review stale skills (${VERSION})" \ --body "$(cat <<'PREOF' ## Stale Skills Detected The following skills may need updates after the latest release: ${{ steps.summary.outputs.summary }} --- ### Update Prompt Paste this into your coding agent (Claude Code, Cursor, etc.): ~~~ ${{ steps.summary.outputs.prompt }} ~~~ PREOF )" \ --head "$BRANCH" \ --base main
What changed
- Run on Latchkey managed runners with one line (
runs-on), which apply the fixes below automatically and self-heal transient failures. This example useslatchkey-small; pick the runner size that fits the job. - Cache dependency installs on the setup step so they are served from cache.
- Add a job timeout so a hung step cannot burn hours of runner time.
What Latchkey heals here
This workflow has steps that commonly fail on transient issues (network, registries, flaky browsers). On Latchkey managed runners they are detected, retried, and self-healed instead of failing your build:
- Dependency installs
This workflow runs 1 job per trigger. On Latchkey the same minutes cost up to 58% less than GitHub-hosted, with zero queue time.