agentflows.sh

Copy-paste GitHub workflows for coding agents.

AgentFlows is a gallery of GitHub Actions for Claude Code, Codex, Cursor, and GitHub Agentic Workflows. Each entry includes the YAML, the trigger, an estimated cost, and a plain record of what was checked against vendor docs — and what has not been executed.

Browse workflows Submit a workflow

20
Workflows: 4
Claude, Codex, Cursor, gh-aw: Est.
Cost, not measured: 20
Not executed yet

Agent

All20 Claude11 Codex3 Cursor2 gh-aw3 Multi-agent1

Trigger

All20 Assignment1 CI failure1 Comment1 Issue opened3 Label6 Pull request6 Schedule2

20 workflows

Claude01
Claude auto PR review
Copied from the official Solutions Guide basic example (prompt shortened).
pull_request: opened, synchronize
Estimated cost~$0.18
Not executed
Checked Oct 16, 2026
Claude02
Claude PR review with progress tracking
`track_progress: true` posts a checkbox tracking comment (per Solutions Guide).
pull_request: opened, synchronize, ready_for_review, reopened
Estimated cost~$0.22
Not executed
Checked Oct 16, 2026
Claude03
@claude mention assistant on issues and PRs
No `prompt` → the action runs in tag/interactive mode and responds to the comment. `trigger_phrase` default is `@claude` (usage.md).
issue_comment / pull_request_review_comment containing @claude
Estimated cost~$0.27
Not executed· 1 caveat
Checked Oct 16, 2026
Claude04
Implement an issue when a label is applied
`label_trigger`: 'The label name that triggers the action when applied to an issue' (usage.md).
issues: labeled (label_trigger)
Estimated cost~$0.69
Not executed· 1 caveat
Checked Oct 16, 2026
Claude05
Implement when an issue is assigned to a bot user
`assignee_trigger`: username that triggers on issue assignment.
issues: assigned (assignee_trigger)
Estimated cost~$0.69
Not executed· 1 caveat
Checked Oct 16, 2026
Claude06
Auto-triage and label new issues
Requires `scripts/gh.sh` and `scripts/edit-issue-labels.sh` copied from the action repo's scripts/ dir (per Solutions Guide).
issues: opened
Estimated cost~$0.06
Not executed· 1 caveat
Checked Oct 16, 2026
Claude07
Review only when critical paths change
Pattern from Solutions Guide 'Review Only Specific File Paths'; paths are examples.
pull_request with paths filter
Estimated cost~$0.18
Not executed· 1 caveat
Checked Oct 16, 2026
Claude08
OWASP-focused security PR review
`use_sticky_comment` keeps one updatable comment (usage.md).
pull_request: opened, synchronize
Estimated cost~$0.22
Not executed
Checked Oct 16, 2026
Claude09
Weekly repo health check
Based on Solutions Guide 'Scheduled Repository Maintenance'.
schedule (cron) + workflow_dispatch
Estimated cost~$0.44
Not executed· 1 caveat
Checked Oct 16, 2026
Claude10
Update docs when API code changes
Based on Solutions Guide 'Documentation Sync on API Changes'.
pull_request with paths: src/api/**
Estimated cost~$0.35
Not executed· 1 caveat
Checked Oct 16, 2026
Claude11
Read failing CI logs and suggest a fix
`additional_permissions` 'Currently supports actions: read' (usage.md).
workflow_run: completed (failure)
Estimated cost~$0.35
Not executed· 1 caveat
Checked Oct 16, 2026
Codex12
Codex PR review posted as a comment
Two-job split keeps the API key job read-only (README example pattern). `:read-only` profile per README Permission profiles.
pull_request: opened
Estimated cost~$0.03 runner + OpenAI tokens (UNVERIFIED)
Not executed· 2 caveats
Checked Oct 16, 2026
Codex13
Codex implements an issue on label
`:workspace` profile lets Codex edit the checkout (README). PR creation is plain git/gh, not an action input.
issues: labeled == codex
Estimated cost~$0.09 runner + OpenAI tokens (UNVERIFIED)
Not executed· 1 caveat
Checked Oct 16, 2026
Codex14
Codex triage with JSON output schema
`output-schema` is forwarded to `codex exec --output-schema` (README inputs table).
issues: opened
Estimated cost~$0.01 runner + OpenAI tokens (UNVERIFIED)
Not executed· 2 caveats
Checked Oct 16, 2026
Cursor15
Launch a Cursor Cloud Agent when an issue is labeled
Fields `prompt.text`, `repos[].url`, `repos[].startingRef`, `autoCreatePR` from the v1 Create An Agent docs. API is public beta.
issues: labeled == cursor
Estimated cost~$0.01 runner + Cursor plan usage (UNVERIFIED)
Not executed· 1 caveat
Checked Oct 16, 2026
Cursor16
Cursor agent drafts a plan first (plan mode)
`mode: plan` 'explores and drafts a plan before coding' (v1 docs).
issues: labeled == planning
Estimated cost~$0.01 runner + Cursor plan usage (UNVERIFIED)
Not executed· 1 caveat
Checked Oct 16, 2026
gh-aw17
gh-aw: agentic issue triage with safe label output
gh-aw source is Markdown + YAML frontmatter compiled with `gh aw compile` to a .lock.yml. Writes go through `safe-outputs`.
issues: opened
Estimated cost~$0.02 runner + engine tokens/credits (UNVERIFIED)
Not executed· 2 caveats
Checked Oct 16, 2026
gh-aw18
gh-aw: run an AI review when `ai-review` label is added
`label_command` removes the label after activation by default (`remove_label: true`).
label_command: ai-review
Estimated cost~$0.03 runner + engine tokens/credits (UNVERIFIED)
Not executed· 1 caveat
Checked Oct 16, 2026
gh-aw19
gh-aw: daily CI optimization coach (GitHub's own)
Trimmed from gh-aw's own .github/workflows/ci-coach.md (removed imports, experiments, AI-credit caps).
schedule: daily + workflow_dispatch
Estimated cost~$0.12 runner + engine tokens/credits (UNVERIFIED)
Not executed· 1 caveat
Checked Oct 16, 2026
Multi-agent20
Label-driven dev loop: triage → plan → implement → red-team → human merge
Mirrors Greg's Trailside Dev Loop labels. Different vendors for author vs red-team reduce shared blind spots. Human gates: adding `ready` approves the plan; merge is manual.
issues/pull_request: labeled
Estimated cost~$0.40–$2.50 per issue end to end (ASSUMPTION)
Not executed· 3 caveats
Checked Oct 16, 2026

Weekly: new and changed flows

New workflows, action version bumps, and measured run costs when a workflow has actually been executed. One email a week.