The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by ingesting and acting upon content from external repository files and user-supplied ideas.\n
Ingestion points: The skill reads project-specific documentation from paths such as docs/specs/*/README.md, docs/project-config.json, and docs/project-reference/.\n
Boundary markers: The instructions lack explicit boundary markers or directives to ignore instructions that might be embedded within the ingested documentation files.\n
Capability inventory: The skill is authorized to use the spawn_agent sub-agent and performs file system operations including Glob discovery and writing artifacts to the team-artifacts/ directory.\n
Sanitization: Technical sanitization or validation of the external content is absent, although the skill includes a cognitive instruction for the AI to remain skeptical and apply critical thinking.

product-owner