find-unplanned-issues
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The bundled TypeScript scripts (
find-unplanned-issues.tsandprune-closed-masterplan-issues.ts) execute shell commands using the GitHub CLI (gh) viaBun.spawnandBun.spawnSyncto list, view, and edit issues. - [EXTERNAL_DOWNLOADS]: The skill fetches issue metadata and hierarchy information from GitHub. The default repository targets official infrastructure managed by the skill author (remotion-dev).
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through external data processing.
- Ingestion points: Issue titles and metadata are fetched from external GitHub repositories using the
gh issue listandgh issue viewcommands in both scripts. - Boundary markers: The output of the audit scripts is presented to the agent in plain text or JSON format without explicit delimiters or instructions to ignore embedded commands within issue titles.
- Capability inventory: The skill possesses the capability to modify live GitHub state (reparenting issues) using
gh issue editinscripts/prune-closed-masterplan-issues.ts. - Sanitization: The scripts do not perform sanitization, filtering, or escaping on the issue titles retrieved from the remote API before displaying them.
Audit Metadata