The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes project-specific documentation such as CONTEXT.md and Architecture Decision Records (ADRs) to sharpen domain models. An attacker who can influence these files could potentially inject instructions that the agent might follow during the domain modeling session.
Ingestion points: CONTEXT.md, CONTEXT-MAP.md, and files within docs/adr/ as referenced in SKILL.md.
Boundary markers: No specific boundary markers or instructions to ignore embedded directives are provided for the content read from these files.
Capability inventory: The skill reads file content for cross-referencing with code and writes updates to documentation files.
Sanitization: No sanitization or validation of the ingested file content is specified.

domain-modeling