arc-bug-fixer

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through the ingestion of untrusted data from bug tracking systems (GitHub, GitLab, Linear).
  • Ingestion points: Bug ticket content (Summary, Steps to Reproduce, Root Cause) fetched via gh, glab, or Linear MCP.
  • Boundary markers: None explicitly mentioned to separate ticket content from agent instructions or to ignore instructions embedded in tickets.
  • Capability inventory: The skill possesses extensive capabilities including file modification, command execution (pnpm, sf, git), and interaction with remote APIs (gh, glab, Linear).
  • Sanitization: No explicit sanitization or validation of the ticket content is mentioned before it is processed by the agent.
  • [COMMAND_EXECUTION]: The skill executes various CLI tools and scripts to perform its tasks.
  • Evidence: Usage of gh, glab, git, pnpm, and Salesforce sf CLI for ticket management, code exploration, testing, and deployment.
  • Evidence: Reference to bin/run.sh within the arc-git-pr-check dependency for shipping code changes.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 03:39 AM
Security Audit — agent-trust-hub — arc-bug-fixer