The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple external sources, creating a surface for indirect prompt injection attacks.
Ingestion points: The skill ingests data from GitHub issue bodies and titles, external plan files, and raw user prompts (SKILL.md, Step 2).
Boundary markers: There are no instructions defining the use of delimiters or boundary markers to distinguish instructions from data when processing external content.
Capability inventory: The skill possesses high-autonomy capabilities, including writing to the project filesystem, performing automated git commits, and executing project-defined scripts (SKILL.md, Procedure and Commit Policy).
Sanitization: The instructions do not specify any sanitization, filtering, or validation steps for the content retrieved from external sources before it is processed by the agent.
[COMMAND_EXECUTION]: The skill uses shell commands to manage the implementation workflow and verify changes.
Version Control: It executes git status and git commit to manage the working tree and persist changes (SKILL.md, Dirty worktree handling and Commit Policy).
Project Scripts: The agent is instructed to run project-defined gates such as lint, format, check, or test scripts prior to committing code (SKILL.md, Commit Policy).

implement