The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection as it ingests untrusted data and has significant system capabilities.
Ingestion points: The skill processes the full conversation history and codebase content discovered via Read, Grep, and Glob (SKILL.md).
Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore or isolate embedded instructions within the ingested data.
Capability inventory: The agent is granted Bash, Write, and Edit tools, allowing for shell execution and modification of the codebase based on the synthesized spec.
Sanitization: No sanitization, filtering, or validation steps are defined for the data before it is synthesized into a specification or published to an issue tracker.
[COMMAND_EXECUTION]: The skill is configured with Bash access in the allowed-tools metadata, which allows for the execution of arbitrary shell commands within the agent's environment.
[DYNAMIC_EXECUTION]: The skill references a local validator script (.agent-skills/skill-standardization/scripts/validate_skill.sh), which constitutes the dynamic execution of project-local script files.

to-spec