code-conventions

Pass

Audited by Gen Agent Trust Hub on Sep 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project documentation (such as AGENTS.md, DESIGN.md, and WORKFLOW.md) and source code to detect convention drift. This ingestion of untrusted external data represents a surface for indirect prompt injection.
  • Ingestion points: Defined in SKILL.md (Procedure steps 1 and 3) and references/drift-detection.md.
  • Boundary markers: Absent; no specific instructions are provided to the agent to treat audited file content as data rather than instructions.
  • Capability inventory: The skill can execute shell commands for linting, testing, and formatting (e.g., go test, go vet, and npm scripts as noted in references/go.md and references/node.md).
  • Sanitization: Absent; the skill does not explicitly filter or sanitize the content of the files it reads.
  • [COMMAND_EXECUTION]: The skill is instructed to run local development and auditing tools.
  • Evidence: SKILL.md requires the agent to "Run available lint, typecheck, test, and formatting checks." references/go.md specifically mentions running "go test" and "go vet" to verify conventions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 22, 2026, 06:23 PM
Security Audit — agent-trust-hub — code-conventions