The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided writing samples (drafts, fragments) to perform voice analysis and style coaching. These samples could contain instructions intended to override the agent's current task.
Ingestion points: Step 1 (Gather Samples) instructs the agent to read writing samples from the user or the project environment (SKILL.md).
Boundary markers: The instructions lack explicit delimitation or directives to treat the writing samples strictly as data rather than instructions.
Capability inventory: The skill utilizes Read, Grep, Glob, Write, and Edit tools, allowing it to modify files (specifically ./docs/voice-profile.md as noted in Step 3).
Sanitization: No sanitization or safety checks are implemented for the content of the processed samples.

cw-voice