maintainable-typescript

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes shell scripts such as scripts/audit-typescript-repo.sh that execute binaries found in the local node_modules/.bin directory (e.g., tsc, oxlint, fallow). These commands are used to audit code health and are scoped to the project directory provided by the user.
  • [EXTERNAL_DOWNLOADS]: The documentation references the fallow tool and companion skills from the fallow-rs organization on GitHub. These are standard references for the recommended development toolchain and do not involve executing untrusted remote scripts.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project source code to perform auditing and refactoring. Mandatory evidence: 1. Ingestion points: Local repository files identified in SKILL.md. 2. Boundary markers: The skill provides extensive doctrine files to guide agent reasoning. 3. Capability inventory: Executes audit scripts and performs file modifications. 4. Sanitization: Uses resolve_target_dir in scripts/lib/common.sh to validate the target path. Severity is safe as this is the primary intended function.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 12:24 PM
Security Audit — agent-trust-hub — maintainable-typescript