checkout
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes
bun iandbun run build. These commands are standard for project setup but involve executing build scripts defined in the repository. - [EXTERNAL_DOWNLOADS]: Uses
bun ito download dependencies from the public NPM registry, which is a common but external operation. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection by processing external data from GitHub PR titles.
- Ingestion points: PR title retrieved via
gh pr view --json titleinSKILL.md. - Boundary markers: Absent; the title is used directly to rename the task.
- Capability inventory: Subprocess execution via
buncommands inSKILL.md. - Sanitization: None; the PR title is not escaped or validated before being used to set the thread title.
Audit Metadata