The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon instructions and validation outputs from the local repository, creating a potential attack surface for indirect prompt injection.\n- Ingestion points: The skill reads repository-specific instructions and Git metadata (status, branch state) as described in SKILL.md.\n- Boundary markers: The skill does not define specific delimiters or instructions for the agent to distinguish repository content from its core operational logic.\n- Capability inventory: The skill is capable of branch management, executing local validation scripts, pushing code to remote repositories, and invoking other installed skills.\n- Sanitization: No sanitization or validation of the repository-provided instructions or script outputs is performed.

apollousa