The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project data and workflow files which could potentially contain malicious instructions.
Ingestion points: The skill reads from ~/.claude/gsd-core/workflows/stats.md and standard project files (git metrics, plans, requirements).
Boundary markers: Absent. There are no instructions to the agent to delimit or ignore instructions embedded in the project data.
Capability inventory: The skill is configured to use the Bash and Read tools.
Sanitization: Absent. The skill does not specify any validation or filtering of the ingested project metrics or workflow content.

gsd-stats