The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the processing of the .agents/docs/orientation.md file. It instructs the agent to read this file and follow the 'Suggested exercise sequence' section, which could be exploited by a malicious repository owner to inject hidden instructions into the agent's context.
Ingestion point: .agents/docs/orientation.md
Boundary markers: Absent. The skill directs the agent to execute the sequence found in the file without explicit delimiters or safety warnings for the content.
Capability inventory: The skill operates in a context where the agent has access to powerful tools like bash and file system modification to facilitate learning exercises.
Sanitization: Absent. There is no validation or filtering of the content retrieved from the orientation file before it influences agent behavior.

learning-opportunities