ts-linter
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPERSISTENCEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of numerous devDependencies from the official npm registry, such as 'eslint', 'typescript', 'typescript-eslint', and various plugins like 'eslint-plugin-unicorn' and 'eslint-plugin-sonarjs'. It also recommends global installation of the 'typescript-language-server'. These are standard, well-known packages used in TypeScript development.
- [COMMAND_EXECUTION]: The skill executes various shell commands for project detection, dependency installation, and linting. It dynamically constructs install and lint commands by reading 'package.json' and CI configuration files (e.g., '.github/workflows/*.yml'). This behavior is necessary for its stated purpose of automating linter setup.
- [PERSISTENCE]: In Claude Code environments, the skill installs 'PostToolUse' and 'Stop' hooks by creating a '.claude/' directory and a 'hooks/lint-typecheck.sh' script. These hooks are configured in '.claude/settings.json' to ensure that linting and type-checking scripts are executed automatically on subsequent file edits and at the end of agent responses.
- [DYNAMIC_EXECUTION]: The skill generates and executes shell scripts ('lint-typecheck-hook.sh') and constructs command strings for execution (e.g., 'CI_LINT_CMD') from project metadata. This is standard behavior for developer tools that automate build and lint workflows.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project metadata which could theoretically be manipulated to influence execution.
- Ingestion points: Reads 'package.json' via 'scripts/detect-project.sh' and CI configuration files in 'SKILL.md' Step 1.
- Boundary markers: No explicit boundary markers or 'ignore embedded instructions' warnings are used when reading these files.
- Capability inventory: The skill can install packages, write files, and execute shell commands.
- Sanitization: Uses 'JSON.parse' in 'detect-project.sh' for structured parsing, but lacks rigorous validation or escaping for extracted values used in command construction.
Audit Metadata