The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external documentation files which creates a potential surface for indirect prompt injection attacks.
Ingestion points: The skill reads SPEC.md, validation.md, background.md, and files within the supporting/ directory from the local environment.
Boundary markers: There are no defined delimiters or specific instructions to treat the file content as non-executable data, increasing the risk that embedded instructions in the docs could influence the agent.
Capability inventory: The agent has the ability to read files, write the resulting executive-summary.md to the local filesystem, and invoke the exec-summary skill.
Sanitization: No sanitization or input validation is performed on the content of the documents being reviewed.
[SAFE]: The skill primarily facilitates a standard analytical workflow. Analysis of the instructions confirms there are no instances of prompt injection, obfuscation, hardcoded credentials, unauthorized network communication, or persistence mechanisms. All operations are local and aligned with the skill's stated purpose.

plan-review