Converts ChatCut handoff manifests to Final Cut Pro XML (FCPXML) format with validation
Both tools have detailed descriptions (100+ chars each) explaining their domain-specific FCPXML operations. However, critical gaps undermine usability: input schemas are present but lack proper JSON Schema type declarations (parameters are documented as strings with no formal constraints), parameter descriptions are minimal or absent, output schemas are completely undocumented, and error handling provides no recovery guidance. The tools are narrowly specialized (FCPXML authoring/validation) with no composition support. Tool names are verb-noun but descriptions are overly technical for LLM selection without explaining WHEN to use each tool vs alternatives.
Author a complete FCPXML from a manifest.json. Encodes patterns verified against a real Final Cut Pro import: per-asset truthful formats, gap-anchored spine with signed lanes, conform=fit, static & keyframed transform scale (keyframe times in SOURCE-time coordinates), cover-rect full-bleed geometry, position in canvas px -> FCP units (canvasH/100), adjust-volume with DTD-legal nested fades, iTT captions with overlap clamping, dedicated library bundle.
Perform structural checks on an FCPXML handoff without claiming FCP import parity. Validates resource IDs, references, media URIs, sequence spine presence, and Final Cut Pro DTD compliance.
Input schemas lack JSON Schema type declarations and constraints. Parameters documented as strings with no enum, pattern, minLength, or format constraints. LLMs cannot validate inputs before calling tools.
Output schemas completely undocumented. No description of what author_fcpxml returns (success/failure? file path? validation errors?) or what validate_fcpxml returns (list of errors? pass/fail boolean?). LLMs cannot plan downstream actions.
Parameter descriptions are minimal or absent. 'manifest_path' and 'output_path' have basic descriptions but no format guidance (absolute vs relative? must exist? directory must be writable?). 'fcpxml' parameter has no description at all.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 42 | 2026-07-28+ | v2 |
No error handling guidance. If manifest.json is malformed, missing assets, or has invalid frame rates, what does the tool return? How should the LLM recover? Source code shows fallback behavior (e.g., 'WARNING asset...has_video but no frame_rate; using project fps') but this is not documented in the tool description.
Tool descriptions are domain-expert focused (mentions 'gap-anchored spine', 'conform=fit', 'DTD-legal nested fades') but lack LLM-friendly context: WHEN should an agent call author_fcpxml vs validate_fcpxml? What is the typical workflow? What does 'truthful formats' mean to an LLM?