mode-xray
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from the local codebase, which is an untrusted source of information.
- Ingestion points: Local project files, architecture definitions, dependency lists, and environment requirement files as specified in the instructions.
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard embedded commands or prompts within the files being scanned.
- Capability inventory: The skill performs filesystem reads and command execution to verify project readiness.
- Sanitization: The skill contains a manual guardrail to "Exclude secrets from documentation," but lacks technical sanitization to prevent adversarial instructions in the codebase from influencing the agent's behavior.
- [COMMAND_EXECUTION]: The guardrails instruct the agent to "Verify quick-start commands are runnable," which necessitates the execution of shell commands discovered within the local project environment.
Audit Metadata