ask-codex
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the local codex CLI tool to perform coding tasks. This is limited by the allowed-tools configuration in the frontmatter to only allow execution of the codex command.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user queries and potentially local code context through an external CLI tool.
- Ingestion points: User prompts are passed as arguments to the CLI in SKILL.md.
- Boundary markers: No delimiters are explicitly used in the provided examples to separate instructions from data.
- Capability inventory: The codex CLI supports autonomous code modification via the --full-auto flag.
- Sanitization: No explicit sanitization or filtering of input is described in the instructions.
Audit Metadata