implement
Pass
Audited by Gen Agent Trust Hub on Apr 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs extensive shell command execution for managing git worktrees, running test suites, linting code, and interacting with the GitHub CLI (
gh). It also executes a project-local script${ARC_ROOT}/scripts/cleanup-orphaned-agents.shfor process management. - [EXTERNAL_DOWNLOADS]: Executes package managers (
pnpm,npm,yarn) to install project dependencies from external registries. - [PROMPT_INJECTION]: The skill demonstrates an indirect prompt injection surface as it processes external documentation and configuration to drive agent behavior.
- Ingestion points: Reads implementation plans (
docs/arc/plans/), design documents,package.json, and.ruler/coding rules. - Boundary markers: None identified for delimiting ingested content during task execution.
- Capability inventory: Full shell execution, git/gh CLI usage, and the ability to spawn specialized build agents.
- Sanitization: No input validation or escaping mechanisms are described for the ingested data before it influences agent operations.
Audit Metadata