implement-spec-with-codex

Pass

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes user-supplied specifications to drive automated code implementation passes, which creates an indirect prompt injection surface.
  • Ingestion points: External specifications provided to the implement-spec workflow (SKILL.md).
  • Boundary markers: The instructions do not specify explicit delimiters or "ignore previous instruction" markers for the input data.
  • Capability inventory: Code execution via delegation to codex exec, file system modifications through diff integration, and version control operations (commits).
  • Sanitization: The skill does not perform automated sanitization of inputs; instead, it enforces security through mandatory manual diff reviews and independent execution verification gates.
  • [COMMAND_EXECUTION]: The skill orchestrates implementation passes using the codex exec capability. This involves generating, integrating, and reviewing code diffs within the local repository environment, including the use of separate worktrees for concurrent implementation tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 24, 2026, 11:36 PM
Security Audit — agent-trust-hub — implement-spec-with-codex