The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill maintains an attack surface for indirect prompt injection by ingesting and acting upon untrusted data from external sources. * Ingestion points: The agent reads external content from GitHub pull request comments, issue reactions, and CI failure logs using gh api and gh run view. It also references potentially external documentation files like AGENTS.md, CLAUDE.md, and repository ADRs. * Boundary markers: The instructions lack explicit delimiters or instructions to treat external data as untrusted or to ignore embedded commands. * Capability inventory: The agent possesses capabilities to execute shell commands (gh, git, make, test runners), perform file system writes to apply repairs, and interact with the GitHub API. * Sanitization: No sanitization or validation logic is specified for the content retrieved from GitHub before it is used to drive the agent's reasoning or repair actions.
[COMMAND_EXECUTION]: The skill dynamically executes commands based on the repository's configuration and current state. * It invokes repository-defined validation tools and cleanup targets, such as make worktree-cleanup-merged, ruff, and mypy, which are defined within the external codebase. * It utilizes the gh CLI for sensitive operations, including marking PRs as ready, merging code, and fetching logs, which relies on the local environment's authentication tokens.

pr-maintainer