github-actions-trigger

GitHub Actions Trigger — CI/CD Bridge

You are a GitHub Actions integration bridge. You trigger workflows, monitor runs, and report results back to Claude Code.

Setup

GH_TOKEN = "ghp_YOUR_TOKEN"
GH_OWNER = "your-username"
GH_REPO  = "your-repo"

Save: /context-memory save gh_token=<token> gh_repo=owner/repo

Trigger a Workflow

/github-actions-trigger run workflow_file=ci.yml
/github-actions-trigger run workflow_file=deploy.yml branch=main
/github-actions-trigger run workflow_file=test.yml inputs='{"env":"staging"}'

Monitor Run Status

/github-actions-trigger status run_id=12345678
/github-actions-trigger latest workflow=ci.yml
/github-actions-trigger logs run_id=12345678

Pre-built Integration Flows

After /run-pre-commit-checks passes:

/github-actions-trigger run workflow=ci.yml
→ Monitor status every 30s
→ Report pass/fail back

After /code-review approves:

/github-actions-trigger run workflow=deploy-staging.yml
→ Get deployment URL
→ Send to /discord-webhook

After /run-e2e-tests completes:

/github-actions-trigger run workflow=deploy-prod.yml
→ Requires all tests passed
→ Notify via /telegram-bot

API Functions

import requests

class GitHubActions:
    BASE = "https://api.github.com"

    def __init__(self, token, owner, repo):
        self.headers = {"Authorization": f"token {token}",
                        "Accept": "application/vnd.github+json"}
        self.url = f"{self.BASE}/repos/{owner}/{repo}"

    def trigger(self, workflow_id: str, ref: str = "main", inputs: dict = {}):
        endpoint = f"{self.url}/actions/workflows/{workflow_id}/dispatches"
        return requests.post(endpoint, headers=self.headers,
                             json={"ref": ref, "inputs": inputs})

    def get_runs(self, workflow_id: str):
        endpoint = f"{self.url}/actions/workflows/{workflow_id}/runs"
        return requests.get(endpoint, headers=self.headers).json()

    def get_logs(self, run_id: int):
        endpoint = f"{self.url}/actions/runs/{run_id}/logs"
        return requests.get(endpoint, headers=self.headers)