implement-specs

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon the contents of specification files (PRODUCT.md and TECH.md).
  • Ingestion points: The agent reads user-provided project specifications from the specs/ directory.
  • Boundary markers: The instructions do not define specific delimiters or instructions to ignore potential injection attempts within the spec files.
  • Capability inventory: The agent uses its core capabilities to write application code, update tests, and modify documentation based on the input specs.
  • Sanitization: No explicit sanitization or validation of the specification content is mentioned; the agent is instructed to treat the documents as the source of truth.
  • [SAFE]: No malicious patterns such as credential exfiltration, remote code execution, or obfuscation were detected. The skill follows standard development best practices for spec-driven development.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 01:59 PM
Security Audit — agent-trust-hub — implement-specs