beo-debug
Pass
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security vulnerabilities were detected in the analyzed files.
- [PROMPT_INJECTION]: The skill instructions provide explicit restrictions ('Hard stops' and 'FORBIDDEN in debug output') that prevent the agent from generating executable patches or bypassing established workflow boundaries. These act as safety reinforcements.
- [COMMAND_EXECUTION]: The skill defines dependencies on local CLI tools (
beads-cli,beads-viewer) for diagnostic tasks. These are referenced as system commands (br,bv) used within a restricted operational scope described as 'display-only'. - [DATA_EXFILTRATION]: There are no indicators of unauthorized data access or network-based exfiltration. All data processing is confined to diagnostic state fields and referenced artifacts within the local environment.
Audit Metadata