standards-directory

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill reads various project design files to construct a prompt for a subagent, which creates an attack surface for indirect prompt injection.
  • Ingestion points: Data is ingested from docs/product/business-flow.md, docs/specs/data/, docs/specs/API, and docs/standards/tech-stack-rule.md.
  • Boundary markers: While templates use markdown headers to delineate sections, there are no explicit delimiters or system instructions for the subagent to ignore embedded commands within the ingested content.
  • Capability inventory: The subagent has read access to the local project files, while the main process handles the writing of generated documentation to the docs/ directory.
  • Sanitization: No sanitization, validation, or escaping is performed on the content of the read files before they are interpolated into the prompt templates.
  • [DYNAMIC_EXECUTION]: The skill dynamically generates a subagent's instructions by reading local template files and performing string replacement on placeholders, effectively assembling executable instructions at runtime based on external file content and user choices.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 02:49 PM
Security Audit — agent-trust-hub — standards-directory