monarch-initiative/mechmaker/github-workflows
- community
Checking that the SKILL.md file is still in the repository.
Turn on, configure, adapt or debug the GitHub workflows of IngestMech: the QC check, the scheduled sweep, and the agent workflows (review, triage, dedupe, PR shepherd, curation scanner, literature scan, compliance, post-review). Use when asked to set up automation, enable an agent, set secrets, change a schedule or a model, tune a prompt, or find out why a workflow failed.
Install
npx vskill@latest install monarch-initiative/mechmaker/github-workflowsSKILL.md
GitHub workflows
The catalog is docs/WORKFLOWS.md: every workflow mechmaker knows, on or off, with what it needs. Read it first.
Turning a workflow on or off
Workflows are Copier answers, so their files stay in step with the template:
List every workflow to keep; the list replaces the old one. Copier writes and deletes the files, and merges your edits to prompts and workflows you kept. Review the diff. To run agent workflows on their schedules, add --data agent_schedules=true.
Do not copy workflow files in by hand. Copier would not know about them.
Setting up an agent workflow
- Credential. A repository secret:
ANTHROPIC_API_KEY(preferred), or
CLAUDE_CODE_OAUTH_TOKEN from claude setup-token. ``bash gh secret set ANTHROPIC_API_KEY ``
- Identity, from the catalog's Identities section:
claude: install the Claude GitHub App.review: create a reviewer App and setMECH_REVIEWER_APP_IDand
MECH_REVIEWER_PRIVATE_KEY, or allow Actions to approve pull requests.
curation-scanner,compliance: create an agent App with contents,
issues and pull requests write, install it, and set MECH_AGENT_APP_ID and MECH_AGENT_PRIVATE_KEY.
- Labels.
just labelsonce. Triage, the scanner and the literature
scan rely on them.
- First run by hand, from the Actions tab or
gh workflow run <file> -f dry_run=true where there is a dry run. Read the run summary before letting it run on a schedule.
Creating Apps, setting secrets and starting workflows act on the live repository. Ask the person before doing any of them.
Adapting to the domain
- Prompts and models come with the first agent workflow: turning one on
adds .github/prompts/ and .github/agent-config.yaml.
- Literature scan is off for this Mech. Turning on the
literature-scan
workflow adds its terms file, conf/literature_scan.yaml, and its recipe.
- Compliance weights are
WEIGHTSinsrc/ingestmech/compliance.py.
Changes to prompts take effect from the default branch, so merge them before judging a run.
When a workflow fails
For qc.yaml, run the failing recipe locally: just qc, or just validate-changed <files>. A term check that says the service is unreachable is an outage, not a data error; re-run later.
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingexample/ingestmech/.claude/skills/github-workflows/SKILL.md
Popularity
- Repository stars
- 3
- 7-day trend score
- 12
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/monarch-initiative/mechmaker/github-workflows)