
from opensoul16
Orchestrates fetching GitHub issues, confirming targets, and spawning code-fix sub-agents to create PRs and handle review comments via the GitHub REST API.
A comprehensive orchestrator skill that fetches issues from a repo, presents them for confirmation, performs pre-flight checks (auth, remotes, existing PRs), and spawns focused sub-agents to implement fixes and open pull requests. It also monitors fix/issue-* PRs for review comments and can spawn review-fix agents to address feedback. The skill uses curl + GitHub REST API and is designed for automated or cron-driven runs.
Use this skill to automate repetitive issue triage and fix workflows: bulk-fix simple issues, run cron-mode processing to pick the next eligible issue, or run review-only checks to handle reviewer comments. It's suited for repositories where small, well-scoped fixes can be auto-implemented by agents.
Designed for autonomous agent runtimes that can call the GitHub API and spawn sub-agents (OpenClaw sessions, ACP/ClawFlow). Works with agent frameworks that support exec/curl and sessions_spawn functionality.
GitHub issue orchestrator that fetches issues, spawns parallel sub-agents to auto-fix them, opens PRs, and handles review comments via GitHub REST API (no gh CLI). Comprehensive 6-phase workflow with cron, watch, fork, and dry-run modes. No bundled scripts — all execution is agent-driven via exec/curl. Token handling embeds GH_TOKEN in git remote URLs which could leak via git remote -v output.
Well-documented orchestrator skill. Main security concern is token exposure via git remote URLs, which is a known trade-off for token-based git auth in CI. No scripts to test. Monolithic SKILL.md (~500 lines) could benefit from extracting sub-agent prompts to references/ directory. DRY improvements possible — token resolution block repeated 3+ times across sub-agent prompts.