The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface by design.
Ingestion points: The skill reads the project's README file and analyzes the local technology stack to "seed" context (found in SKILL.md).
Boundary markers: Absent. There are no instructions or delimiters provided to ensure the agent ignores potentially malicious instructions embedded within the README or project files being analyzed.
Capability inventory: The skill utilizes file-writing capabilities through the faf sync command, which modifies sensitive agent configuration files like CLAUDE.md and AGENTS.md (found in SKILL.md).
Sanitization: Absent. The skill does not describe any validation or sanitization of the content extracted from the user's project files before incorporating it into the AI's context.
[COMMAND_EXECUTION]: The skill instructs the agent to perform local shell operations.
The instructions direct the agent to run a series of CLI commands: faf auto, faf score, faf go, and faf sync. While these are part of the intended functionality, they involve executing local binaries that must be present on the system.

faf-context