oops
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a process for capturing and analyzing untrusted data from environment diagnostics and command failures. \n
- Ingestion points: Captured error output from reproduction commands and project-level configuration files such as .noodle.toml and .noodle/ state files. \n
- Boundary markers: Absent; the instructions prompt the agent to capture exact output without providing specific delimiters or instructions to ignore embedded commands. \n
- Capability inventory: The agent is directed to execute shell commands, modify local project files, and perform git commits. \n
- Sanitization: The skill does not include steps to sanitize or validate the content of error logs or configuration files before the agent processes them.
Audit Metadata