The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by interpolating untrusted user requirements into generated implementation plans that include executable commands.
Ingestion points: User-provided specifications or requirements used to define the implementation tasks.
Boundary markers: The skill does not define specific delimiters or instructions to ignore instructions embedded within the user's requirements.
Capability inventory: The skill instructs the agent to write files to docs/plans/ and prepare shell commands such as pytest and git for execution by subsequent skills.
Sanitization: No input validation or sanitization is performed on the user requirements before they are incorporated into the plan documents.

writing-plans