framesail

Director note (one scene)

director_note

Edit ONE scene with a natural-language note (the same director chat the editor UI uses): move/restyle/add/remove layers and overlays, retime, etc. Synchronous — returns the applied mutations + updated scene. Address the scene by project_id + segment_number (preferred — always resolves to the current scene) or by scene_id from a fresh list_scenes call. Scene durations are locked to the narration track: don't ask to extend/shorten a scene except the final one — for a held-shot feel mid-video, ask for calmer/static motion on the scene instead. For notes spanning the whole video use project_director_note instead.

Input Schema

TableJSON Schema

Name	Required	Description
`message`	Yes	Natural-language edit note for this scene, e.g. "move the caption to the top" or "remove the overlay"
`scene_id`	No	ID of the scene to edit, from a FRESH list_scenes call — scene ids change whenever segments are edited (split/combine/update-content), so never reuse ids saved earlier; prefer project_id + segment_number
`project_id`	No	Project ID; required (with segment_number) when not passing scene_id
`segment_number`	No	1-based segment number of the scene to edit (from list_scenes or get_segments) — preferred over scene_id because it is resolved to the current scene at call time
`conversation_history`	No	Prior chat turns as [{"role": ..., "content": ...}] to continue an editing conversation on this scene; omit to start fresh

TDQS

A5/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With annotations all false (no positive hints), the description carries full responsibility and does so thoroughly. It discloses synchronous execution, return contents ('applied mutations + updated scene'), locked scene durations, and that segment_number always resolves to the current scene. This goes well beyond the bare schema.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Four dense sentences, each earning its place: purpose/capabilities, synchronous return, addressing/scene-id guidance, and the duration constraint with a sibling-tool pointer. No redundancy, front-loaded with the core action.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 5-parameter tool with no output schema, the description is remarkably complete. It covers the return value (mutations + updated scene), critical constraints (duration lock, fresh scene_id), and the sibling alternative. The only minor gap is unspecified precedence if both scene_id and segment_number are passed, but this is negligible given the explicit preference.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, but the description adds crucial semantic guidance: it explains why project_id + segment_number is preferred (resolves to current scene), warns that scene_ids become stale after edits, and clarifies that conversation_history can continue a prior chat. This is genuinely useful parameter context beyond the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with 'Edit ONE scene with a natural-language note' — a specific verb (edit) and resource (one scene) — and enumerates concrete actions (move/restyle/add/remove layers, retime). It also distinguishes itself from the sibling tool 'project_director_note' by explicitly stating that whole-video notes should go there.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Provides explicit when-to-use guidance: 'For notes spanning the whole video use project_director_note instead.' Also gives scene-addressing best practices (prefer project_id + segment_number, avoid stale scene_ids) and behavioral constraints (don't extend/shorten scenes except the final one). This clearly orients the agent on how and when to invoke the tool.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All framesail tools Inspect

Try in Browser

RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

TDQS

A3.6/5.0

Disambiguation3/5

Several tool families overlap in purpose, such as await_jobs/get_workflow_status/get_pipeline_progress, update_segment_content/update_segment_prompts, director_note/project_director_note, and scan_script/rescan_voice_blocks. The descriptions do a good job distinguishing them, but an agent must read carefully to avoid misselection, and there are more than a couple of confusable pairs.

Naming Consistency4/5

The set overwhelmingly follows a verb_noun snake_case convention with clear prefixes like get_, list_, set_, update_, create_, and delete_. Minor exceptions such as director_note, project_director_note, browse_audio_library, and whoami keep it from being perfectly consistent.

Tool Count1/5

At 72 tools, this is far beyond the 50+ extreme range and creates a heavy navigation burden for an agent. Even though the pipeline is complex, this many tools is not well-scoped for an MCP surface.

Completeness4/5

The surface covers the full script-to-export pipeline: styles, assets, voices, storyboards, segments, scenes, and rendering all have substantial lifecycle support. Some gaps exist—no delete_channel, delete_segment, delete_voice_block, or delete_provider_key—but most missing operations can be worked around through existing tools.

Resources

Need Help?