pr-status-triage
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- Indirect Prompt Injection: The skill processes external data derived from CI logs and GitHub pull request threads, creating a surface for potential instruction injection.\n
- Ingestion points: The agent reads generated markdown files such as
job-{id}.mdandthread-{N}.mdlocated inscripts/pr-status/.\n - Capability inventory: The skill provides the agent with access to shell commands including
gh,node,pnpm, andcargo.\n - Boundary markers: The instructions do not specify the use of delimiters or warnings to ignore embedded instructions within the processed files.\n
- Sanitization: There is no evidence of filtering or sanitization performed on the content before it is read by the agent.\n- Command Execution: The skill automates development workflows by executing local repository scripts and command-line tools.\n
- Evidence: The workflow involves running commands like
node scripts/pr-status.jsandgh run rerun <run-id> --failed.\n - Context: These operations are performed to reproduce CI failures and manage pull request threads, utilizing resources provided by the vendor.
Audit Metadata