The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and apply guidelines from project-specific files, which creates a surface for untrusted instructions to influence agent behavior.
Ingestion points: The agent is directed to read AGENTS.md and .jimmy/constitution.md from the target project repository to determine additional engineering principles.
Boundary markers: The skill does not provide specific delimiters or warnings to ignore potentially malicious instructions embedded within these external project files.
Capability inventory: The skill guides the agent's decision-making process for code changes, test requirements, and documentation storage in directories like .jimmy/adr/ and .jimmy/docs/.
Sanitization: There is no evidence of sanitization or validation logic to filter content read from these external files.
[EXTERNAL_DOWNLOADS]: The documentation includes a reference to a public repository (github.com/xoai/sage) as a source for upstream engineering documentation and optional reading.

constitution