dev-cli-consistency-audit
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from CLI command outputs and project documentation.\n
- Ingestion points: The workflow involves reading the output of CLI
--helpcommands and the contents of README files (SKILL.md, Step 1 and Step 3).\n - Boundary markers: The skill lacks explicit delimiters or instructions for the agent to disregard embedded commands within the ingested data.\n
- Capability inventory: The agent is instructed to execute local commands and modify source code/documentation files based on its findings (SKILL.md, Step 1 and Step 5).\n
- Sanitization: No sanitization or filtering logic is provided for the data read from external sources.\n- [COMMAND_EXECUTION]: The skill requires the agent to execute local binaries to inspect their help interfaces.\n
- Evidence: The workflow (Step 1) explicitly instructs running
<cmd> --helpto inventory commands. While this is the primary purpose of the skill, executing arbitrary binaries carries inherent risk if the binaries are malicious.
Audit Metadata