skills/erphq/skills/coming-soon/Gen Agent Trust Hub

coming-soon

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFECREDENTIALS_UNSAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill directs the agent to obtain authentication credentials from the local path ~/.build-host/credentials.json.
  • [DATA_EXFILTRATION]: The skill transmits generated site content and deployment requests to external build.host API endpoints.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via the processing of untrusted user-provided project briefs.
  • Ingestion points: The user-provided project brief used for content generation in the operating procedure.
  • Boundary markers: No delimiters or safety instructions are defined to isolate the brief content from the agent's instructions.
  • Capability inventory: The agent can write files to the project directory and perform network deployments via the build.host API.
  • Sanitization: No specific sanitization or validation of the brief's content is performed before the site code is generated and shipped.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 04:44 PM
Security Audit — agent-trust-hub — coming-soon