The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of system commands including git (worktrees, logs, commits), gh (issue listing, PR creation, status checks), and pnpm (installation and builds) to orchestrate the workflow.
[EXTERNAL_DOWNLOADS]: The skill performs package installations using pnpm i, which fetches dependencies from the configured registry (typically npmjs.com) for the local project being worked on.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources, including GitHub issue titles/descriptions and the terminal output generated by autonomous sub-agents.
Ingestion points: Data enters the context via gh issue list and by reading the tail output of the spawned mc instances.
Boundary markers: The skill does not define explicit delimiters or boundary markers for the ingested content in its instructions.
Capability inventory: The supervisor agent has the ability to execute shell commands, write report files to the local directory, and create pull requests on GitHub.
Sanitization: There is no automated sanitization mentioned, though the skill explicitly instructs the supervisor to "Review all mc output before creating PRs" and mandates user approval.
[DYNAMIC_EXECUTION]: The skill manages a complex lifecycle of autonomous sub-processes (mc instances) using pnpx tsx, monitors their progress via PIDs, and dynamically generates new prompts based on their diffs and logs to handle restarts and timeouts.

gh-bulk-issues