The Agent Skills Directory

[SAFE]: No security issues were identified in the skill logic or structure.
[NO_CODE]: The skill is composed exclusively of markdown files and does not contain any executable scripts or binaries.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes untrusted input from product briefs.
Ingestion points: Brief files are loaded in Stage 1 from paths like .pawbytes/prodig-suites/products/{product-slug}/brief.md.
Boundary markers: No delimiters are used to wrap external content or instruct the agent to ignore embedded instructions.
Capability inventory: The skill's primary actions are reading and writing markdown files within the project's .pawbytes directory.
Sanitization: No input validation or sanitization is performed on the ingested brief content before it is incorporated into plans and handoff documents.

paw-ps-concept-to-product-plan