bruhs
Pass
Audited by Gen Agent Trust Hub on Feb 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- REMOTE_CODE_EXECUTION (LOW): The skill uses
npxandpnpm createto fetch and execute external packages. This is a standard part of its scaffolding and MCP integration functionality but involves executing code from unverified sources. - PROMPT_INJECTION (LOW): The skill's
/bruhs peepand/bruhs cookcommands ingest external data from GitHub PR comments and Linear tickets. This creates an indirect prompt injection surface where a malicious user could potentially influence agent behavior through these comments. - Ingestion points: GitHub PR comments (
gh api), Linear ticket descriptions. - Boundary markers: Absent. No specific delimiters or warnings are used to isolate untrusted data.
- Capability inventory: The agent has permission to execute shell commands (
git,gh), write to the filesystem, and perform network requests via MCP. - Sanitization: Absent. External text is processed directly within the agent context.
- COMMAND_EXECUTION (SAFE): Extensive use of
gitandghis core to the skill's purpose and appears to be implemented using safe, standard patterns.
Audit Metadata