llms-txt-and-crawler-access
Pass
Audited by Gen Agent Trust Hub on Jul 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's primary logic is contained within natural language instructions in
SKILL.md, which includes explicit guardrails preventing the agent from recommending access to sensitive or private areas like admin panels, carts, or checkout sessions. - [SAFE]: The provided Python and shell scripts in the
eval/directory are designed for offline evaluation and CI/CD testing. Therun-eval.shscript specifically hardens the testing environment by disabling network sockets viasitecustomize.py, demonstrating a security-conscious design for the evaluation process. - [SAFE]: No evidence of data exfiltration, credential exposure, or unauthorized command execution was found across the analyzed files.
Audit Metadata