setup-putior-ci
À propos
Cette compétence Claude configure un pipeline CI/CD GitHub Actions pour régénérer et valider automatiquement les diagrammes de flux de travail putior à chaque envoi. Elle génère le fichier YAML de workflow nécessaire, un script R avec des marqueurs sentinelles pour la génération des diagrammes, et s'intègre aux fichiers README pour des mises à jour en place. Utilisez-la pour garantir que les diagrammes de flux reflètent toujours le code actuel, particulièrement dans les projets multi-contributeurs où la régénération manuelle est sujette aux erreurs.
Installation rapide
Claude Code
Recommandénpx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/setup-putior-ciCopiez et collez cette commande dans Claude Code pour installer cette compétence
Documentation
Set Up putior CI/CD
Configure GitHub Actions to automatically regenerate workflow diagrams when source code changes, keeping documentation in sync with code.
When to Use
- Workflow diagrams should always reflect the current state of the code
- The project has CI/CD and wants automated documentation updates
- Multiple contributors may change workflow-affecting code
- Replacing manual diagram regeneration with automated pipeline
Inputs
- Required: GitHub repository with putior annotations in source files
- Required: Target file for diagram output (e.g.,
README.md,docs/workflow.md) - Optional: putior theme (default:
"github") - Optional: Source directories to scan (default:
"./R/"or"./src/") - Optional: Branch to trigger on (default:
main)
Procedure
Step 1: Create GitHub Actions Workflow
Create the workflow YAML file for automated diagram generation.
# .github/workflows/update-workflow-diagram.yml
name: Update Workflow Diagram
locale: caveman-lite
source_locale: en
source_commit: 82c77053
translator: "Julius Brussee homage — caveman"
translation_date: "2026-04-19"
on:
push:
branches: [main]
paths:
- 'R/**'
- 'src/**'
- 'scripts/**'
permissions:
contents: write
jobs:
update-diagram:
if: github.actor != 'github-actions[bot]'
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- uses: r-lib/actions/setup-r@v2
with:
use-public-rspm: true
- name: Install putior
run: |
install.packages("putior")
shell: Rscript {0}
- name: Generate workflow diagram
run: |
Rscript scripts/generate-workflow-diagram.R
- name: Commit updated diagram
run: |
git config --local user.name "github-actions[bot]"
git config --local user.email "github-actions[bot]@users.noreply.github.com"
git add README.md docs/workflow.md # Adjust to match your target files
git diff --staged --quiet || git commit -m "docs: update workflow diagram [skip ci]"
git push
Expected: File created at .github/workflows/update-workflow-diagram.yml.
On failure: Ensure the .github/workflows/ directory exists. Adjust the paths filter to match where annotated source files live in the repository.
Step 2: Write Generation Script
Create the R script that generates the diagram and updates target files using sentinel markers.
# scripts/generate-workflow-diagram.R
library(putior)
# Scan source files for annotations (exclude build scripts to avoid circular refs)
workflow <- put_merge("./R/", merge_strategy = "supplement",
exclude = c("generate-workflow-diagram\\.R$"),
log_level = NULL) # Set to "DEBUG" to troubleshoot CI diagram generation
# Generate Mermaid code
mermaid_code <- put_diagram(workflow, output = "raw", theme = "github")
# Read target file (e.g., README.md)
readme <- readLines("README.md")
# Find sentinel markers
start_marker <- "<!-- PUTIOR-WORKFLOW-START -->"
end_marker <- "<!-- PUTIOR-WORKFLOW-END -->"
start_idx <- which(readme == start_marker)
end_idx <- which(readme == end_marker)
if (length(start_idx) == 1 && length(end_idx) == 1 && end_idx > start_idx) {
# Replace content between sentinels
new_content <- c(
readme[1:start_idx],
"",
"```mermaid",
mermaid_code,
"```",
"",
readme[end_idx:length(readme)]
)
writeLines(new_content, "README.md")
cat("Updated README.md workflow diagram\n")
} else {
warning("Sentinel markers not found in README.md. Add them manually:\n",
start_marker, "\n", end_marker)
}
# Also write standalone diagram file
writeLines(
c("# Workflow Diagram", "",
"```mermaid", mermaid_code, "```"),
"docs/workflow.md"
)
cat("Updated docs/workflow.md\n")
Expected: Script at scripts/generate-workflow-diagram.R that reads annotations, generates Mermaid code, and replaces content between sentinel markers.
On failure: If put_merge() returns empty, check that source paths match the repository layout. Adjust "./R/" to the actual source directory.
Step 3: Configure Auto-Commit
The workflow must avoid infinite loops where an auto-commit re-triggers the same workflow. Pushes made with the default GITHUB_TOKEN typically do not trigger new workflow runs, but the workflow also includes an explicit if: guard on the job as a safety net.
Key configuration points:
permissions: contents: writegrants push accessif: github.actor != 'github-actions[bot]'skips the job when the push came from the bot itselfgit diff --staged --quiet || git commitonly commits if there are changes[skip ci]in the commit message is a convention some CI systems honor (not built into GitHub Actions, but useful as a signal)- Bot identity used for commits:
github-actions[bot]
Expected: The workflow only commits when diagrams actually change. No empty commits, no infinite loops.
On failure: If push fails with permission denied, check repository settings: Settings > Actions > General > Workflow permissions must be set to "Read and write permissions".
Step 4: Add Sentinel Markers to README
Insert sentinel markers in the target file where the diagram should appear.
## Workflow
<!-- PUTIOR-WORKFLOW-START -->
<!-- This section is auto-generated by putior CI. Do not edit manually. -->
```mermaid
flowchart TD
A["Placeholder — will be replaced on next CI run"]
<!-- PUTIOR-WORKFLOW-END -->
**Expected:** Sentinel markers in README.md (or other target file). The content between them will be replaced on each CI run.
**On failure:** Ensure markers are on their own lines with no leading/trailing whitespace. The script matches exact line content.
### Step 5: Test the Pipeline
Trigger the workflow and verify the diagram updates.
```bash
# Make a small change to trigger the workflow
echo "# test" >> R/some-file.R
git add R/some-file.R
git commit -m "test: trigger workflow diagram update"
git push
# Monitor the GitHub Actions run
gh run watch
# Verify the diagram was updated
git pull
cat README.md | grep -A 5 "PUTIOR-WORKFLOW-START"
Expected: GitHub Actions run completes successfully. The diagram between sentinel markers in README.md is updated with current workflow data.
On failure: Check the Actions log for errors. Common issues:
putiorpackage not available: add toDESCRIPTIONSuggests or install explicitly in the workflow- Source path wrong: the R script's
put_merge()path must be relative to the repo root - No sentinel markers: the script warns but doesn't crash; add markers to README.md
Validation
-
.github/workflows/update-workflow-diagram.ymlexists and is valid YAML -
scripts/generate-workflow-diagram.Rruns without errors locally - README.md contains
<!-- PUTIOR-WORKFLOW-START -->and<!-- PUTIOR-WORKFLOW-END -->sentinels - GitHub Actions workflow triggers on push to the correct branch and paths
- Diagram content between sentinels is updated after a workflow run
- Job-level
if:guard prevents infinite commit loops from bot pushes - No changes = no commit (idempotent)
Common Pitfalls
- Infinite CI loops: Pushes with the default
GITHUB_TOKENtypically don't trigger new runs, but always add an explicitif: github.actor != 'github-actions[bot]'guard on the job. The[skip ci]tag in the commit message is a useful convention but is not a built-in GitHub Actions mechanism. - Permission denied on push: GitHub Actions needs write permission. Set
permissions: contents: writein the workflow file, or configure it in repository settings. - Sentinel marker mismatch: If markers have trailing spaces, leading tabs, or are on the same line as other content, the script won't find them. Keep markers on their own clean lines.
- Source path mismatch: The R script runs from the repo root. Paths like
"./R/"or"./src/"must match the actual directory structure. - Package installation in CI: If the project uses renv, the CI workflow needs
renv::restore()before putior is available. Alternatively, install putior explicitly in the workflow. - Large repos slowing CI: For repos with many source files, limit the
pathstrigger filter to directories that contain PUT annotations, not the entire repo.
Related Skills
generate-workflow-diagram— the manual version of what this CI automatessetup-github-actions-ci— general GitHub Actions CI/CD setup for R packagesbuild-ci-cd-pipeline— broader CI/CD pipeline designannotate-source-files— annotations must exist before CI can generate diagramscommit-changes— understanding auto-commit patterns
Dépôt GitHub
Compétences associées
content-collections
MétaCette compétence propose une configuration éprouvée en production pour Content Collections, un outil axé sur TypeScript qui transforme des fichiers Markdown/MDX en collections de données typées de manière sûre avec une validation Zod. Utilisez-la lors de la création de blogs, de sites de documentation ou d'applications Vite + React riches en contenu pour garantir la sécurité de typage et la validation automatique du contenu. Elle couvre tout, de la configuration du plugin Vite et de la compilation MDX à l'optimisation des déploiements et la validation des schémas.
polymarket
MétaCette compétence permet aux développeurs de créer des applications avec la plateforme de marchés prédictifs Polymarket, incluant l'intégration d'API pour le trading et les données de marché. Elle fournit également une diffusion de données en temps réel via WebSocket pour surveiller les transactions en direct et l'activité du marché. Utilisez-la pour mettre en œuvre des stratégies de trading ou pour créer des outils traitant les mises à jour de marché en direct.
creating-opencode-plugins
MétaCette compétence aide les développeurs à créer des plugins OpenCode qui s'interconnectent avec plus de 25 types d'événements tels que les commandes, les fichiers et les opérations LSP. Elle fournit la structure du plugin, les spécifications de l'API événementielle et les modèles d'implémentation pour les modules JavaScript/TypeScript. Utilisez-la lorsque vous avez besoin d'intercepter, de surveiller ou d'étendre le cycle de vie de l'assistant IA OpenCode avec une logique personnalisée pilotée par les événements.
sglang
MétaSGLang est un framework de service LLM haute performance spécialisé dans la génération rapide et structurée pour les workflows JSON, regex et agentiques grâce à son cache de préfixe RadixAttention. Il offre une inférence nettement plus rapide, particulièrement pour les tâches avec des préfixes répétés, ce qui le rend idéal pour les sorties complexes et structurées ainsi que les conversations multi-tours. Choisissez SGLang plutôt que des alternatives comme vLLM lorsque vous avez besoin d'un décodage contraint ou que vous construisez des applications avec un partage étendu de préfixes.
