building-spec

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow requiring the agent to ingest and follow instructions from external markdown files (BUILDING_SPEC.md, DIMENSIONS.md) found in the project environment.
  • Ingestion points: Reads project-specific configuration files (BUILDING_SPEC.md, DIMENSIONS.md) searched in the project root and parent directories as defined in SKILL.md.
  • Boundary markers: Absent; the instructions do not establish delimiters or guidelines to help the agent distinguish between architectural data and potentially malicious natural language instructions within the specification files.
  • Capability inventory: None; the skill does not define or utilize any high-risk tools such as network requests, shell access, or file-writing operations.
  • Sanitization: Absent; the skill explicitly directs the agent to treat the external file content as "ground truth" for all subsequent generation and validation tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 04:58 AM
Security Audit — agent-trust-hub — building-spec