codex-workflow-recommender
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill implements strict privacy boundaries by explicitly forbidding the inclusion of sensitive diagnostic data such as authentication tokens, provider secrets, or environment variables in its reports as defined in the core instructions and specialized references.
- [EXTERNAL_DOWNLOADS]: The skill references official documentation from a well-known service (learn.chatgpt.com). These references are used for informational purposes and do not involve the execution of remote code or the acquisition of unverifiable dependencies.
- [COMMAND_EXECUTION]: Shell access is strictly constrained in the frontmatter and manifest to specific read-only commands. The skill instructions emphasize a read-only workflow, prohibiting any modification of the local environment or repository state.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests potentially untrusted repository data such as AGENTS.md, it mitigates risk by operating under a strictly read-only contract and providing only evidence-backed recommendations without the capability to execute them or modify files directly.
Audit Metadata