Deterministic structural framing analysis for AI-generated documents with default-on frame-divergence block per FRAME_DIVERGENCE_CONTRACT_v1 c1.0. MCP surface delegates V4.2 judgment to the caller's agent model; zero Frame Check LLM cost per query.
Frame Check MCP defines 2 tools with complete input schemas and reasonable descriptions. Both tools follow verb_noun naming (frame-check, frame-compare) and accept structured parameters with type definitions. Descriptions are adequate (100-150 chars) but lack actionable guidance on when to use each tool vs. the other, and neither specifies output schema structure. Parameters are well-typed (string inputs with descriptions) but lack enums, ranges, or validation constraints. Error handling is not visible in the tool definitions. The server implements prompts and resources features, suggesting broader capability, but tool definitions alone are moderate quality, typical of a domain-specific analysis tool that prioritizes correctness over LLM-friendly composition.
Analyze a document for structural framing patterns, epistemic stance, temporal context, and frame divergence. Returns deterministic measurements with agent guidance on interpretation and citation.
Compare framing patterns across two documents. Identifies frame divergence, shared frames, and epistemic differences between documents.
Output schema not documented. Tools return 'deterministic measurements' and 'frame divergence findings' but no structured schema is visible. LLMs cannot plan downstream calls or extract specific fields without knowing the response structure.
No error handling guidance. Tool descriptions do not explain what happens on invalid input (e.g. empty document, unsupported domain_hint), what errors are retryable, or how to recover. LLMs cannot self-correct on failure.
No tool differentiation guidance. Both tools analyze framing patterns; descriptions do not explain when to call frame-check vs. frame-compare. LLMs must guess based on parameter names alone.
domain_hint parameter lacks enum constraint. Accepts 'news', 'academic', 'social_media', 'policy', 'marketing' but no enum is declared. LLMs may hallucinate invalid values like 'blog' or 'legal'.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 58 | <=2025-11-25 | v2 |
divergence_rendering parameter lacks enum constraint. Accepts 'block', 'summary', 'detailed' but no enum is declared. LLMs may pass invalid values.