A NestJS-based server for analyzing code files with syntax checking, linting, static analysis, and custom code analysis
This server has a single tool, handleFileChange, with critical quality gaps. The tool name violates verb-noun conventions (should be analyze_file or check_file_syntax, not handleFileChange in camelCase for an MCP tool). The description is present (194 chars) but vague about what analysis is performed and when to use it. Input schema exists with proper types (filename: string, content: string) and descriptions, but output schema is completely undocumented, the tool returns a bare string 'File change processed and analyzed' with no structured response. The implementation performs multiple disparate analyses (syntax checking, linting, static analysis, custom analysis including complexity, security, dead code detection, pattern matching) but the tool description does not explain this scope or guide LLM selection. Error handling is minimal, all errors are logged to console with no actionable recovery guidance returned to the agent. The tool is not idempotent (analysis results vary with code content). No pagination, no structured output, no field chaining for downstream tools.
Processes file changes by performing syntax checking, linting, static analysis, and custom analysis including complexity calculation, security vulnerability checks, dead code detection, and pattern matching
Tool name violates verb-noun convention. 'handleFileChange' uses camelCase and generic verb 'handle'. Should be 'analyze_file_syntax', 'check_code_quality', or similar action verb + noun.
Output schema completely undocumented. Tool returns a bare string ('File change processed and analyzed') with no structured fields. LLM cannot extract analysis results (complexity score, vulnerabilities, issues found, etc.) for downstream reasoning or user display.
Tool description lacks clarity on scope and use case. States the tool 'Processes file changes by performing syntax checking, linting, static analysis, and custom analysis' but does not explain WHEN to use it vs other analysis tools, WHAT constitutes 'custom analysis', or WHAT the output contains.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 30 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 16 | - | v1 |
Error handling returns no actionable guidance. Errors are logged to console (console.error for syntax errors, undefined behavior for ESLint/TS compiler errors). Agent receives generic success response even if analysis fails, so it cannot detect or recover from errors.
No input validation or constraint documentation. Parameters 'filename' and 'content' lack format/length constraints. Tool accepts any string, including empty content or invalid filenames, with no clear error messages.
Analysis results are lost. Console logs (complexity, linting issues, static analysis errors, security checks, dead code) are written to stdout/stderr only. None of this data is returned in the response, so the agent and user cannot access the analysis findings.
Multiple disparate analyses crammed into one tool without composition. Tool name and description suggest it handles 'file changes' generically, but implementation performs syntax check, linting, static analysis, complexity calculation, security scanning, dead code detection, and pattern matching, five distinct concerns that should be separate tools or at least documented as sub-operations with their own output fields.