crane-audit

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external GitHub pull requests using gw gh pr view and gw gh pr diff (SKILL.md). This content, which includes code changes and metadata, is processed by the agent to generate a compliance report. There are no explicit boundary markers or sanitization procedures described to isolate these untrusted inputs. Consequently, an attacker could embed malicious instructions within the PR diff (e.g., in code comments or strings) to deceive the agent or manipulate the audit results. The skill possesses the capability to execute shell commands for type checking (SKILL.md).
  • [COMMAND_EXECUTION]: The skill instructions direct the agent to execute shell commands to facilitate the audit workflow. These include gw gh pr view and gw gh pr diff for data gathering, and bun svelte-check, tsc --noEmit, and pnpm run check for static analysis and type checking of the affected codebase (SKILL.md, references/compliance-checks.md). These commands are part of the intended functionality for verifying pull request compliance.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 01:31 AM