Loading ToolBench
A collection of AI skills and tools for design-to-code, database design, secret scanning, debugging, and other development tasks
Three tools with READ_ONLY operations. schema_validator has minimal schema (1 param, no output documented). gitleaks_detect has 7 params with types and descriptions, but output schema undocumented. gitleaks_pipeline has 8 params with types and descriptions, but output schema undocumented. Descriptions are present (60-150 chars) but lack WHEN/WHY guidance. No error handling patterns visible. No tool annotations (readOnlyHint, etc.). Parameter descriptions exist but lack format/constraint details (e.g., what constitutes a valid 'project_path'?). No evidence of idempotency, pagination, or recovery guidance.
Gitleaks secret scanning - Scans git repositories, directories, or specific commits for hardcoded secrets, API keys, tokens, and credentials
Gitleaks Secret Scanning Pipeline - Runs Gitleaks scans, manages baselines, evaluates findings, and generates remediation reports
Database schema validation - Validates Prisma schemas and checks for common issues including syntax, missing relations, index recommendations, and naming conventions
Output schemas not documented for any tool. LLMs cannot plan downstream calls or extract required fields without knowing response structure.
Parameter descriptions lack format/constraint details. E.g., 'project_path' has no guidance on absolute vs relative paths, or validation rules. 'mode' enum is documented but no explanation of when to use 'detect' vs 'protect'.
No error handling guidance. Tools do not indicate what errors are retryable, user-fixable, or fatal. No recovery suggestions (e.g., 'If schema validation fails, check syntax with...').
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 47 | 2026-07-28+ | v2 |
No tool annotations. All three tools are READ_ONLY but lack readOnlyHint annotation. This forces LLMs to infer safety from descriptions alone, increasing misuse risk.
Descriptions lack WHEN/WHY context. E.g., gitleaks_detect says what it does but not when to call it vs gitleaks_pipeline, or what the user should do with results.