The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions create a surface for indirect prompt injection by directing the agent to ingest external, potentially untrusted project data.
Ingestion points: SKILL.md instructs the agent to read CONTEXT.md to align test names and interface vocabulary with the project's domain language.
Boundary markers: The instructions lack explicit delimiters or warnings for the agent to ignore instructions that might be embedded within the project's CONTEXT.md file.
Capability inventory: The skill is designed for an environment where the agent can read files, write code, and potentially execute tests or interact with other design-related skills.
Sanitization: There are no documented steps for validating or filtering the content of the ingested project files.

tdd