The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it instructs the agent to process external data (project information and deployment logs) from the build.host API without specified sanitization or boundary markers.
Ingestion points: GET /api/projects, GET /api/projects/:uuid/logs (defined in SKILL.md).
Boundary markers: None identified in the instructions.
Capability inventory: Executes deployment operations and status checks via API endpoints.
Sanitization: No sanitization or filtering logic is specified for the ingested API data.
[CREDENTIALS_UNSAFE]: The skill identifies ~/.build-host/credentials.json as the source for authentication.
This path is the official configuration location for the build.host platform.
The instructions include mandatory safety rules to prevent the printing or sharing of secrets, which mitigates the risk of credential exposure.

rollback