Project-level info: file path, dirty flag, all items with type/summary, active item. Start here.

Detailed comp info: size, fps, duration, work area, motion blur, layer summaries. Pass a comp name or item id — or an ARRAY of them to fetch several comps in one call.

Full layer info: transform, effects, masks, text, shape contents, keyframes (incl. interpolation names and bezier ease speed/influence). layerIndex accepts one index, an ARRAY of indices, or 'all' — auditing a whole comp is one call. detail:'summary' drops properties still at their default value (unkeyed, no expression) — start there for audits; the full tree is often 100KB+ per comp. Set includeProperties=false to skip the property tree walk entirely.

Render one or more frames to PNG. Use to visually verify edits. The agent's 'eyes' — pair with mutations for a visual feedback loop. Headless and deterministic. Pass time for one frame, or times for several in ONE call (motion checks): files land at _.png and are listed in frames. contactSheet tiles every frame into one labelled PNG (_sheet.png) so a motion check is one image; analyze measures each capture (uniform edge bands = black bars / transparent margins, content bounds, coverage) so off-frame elements and letterboxing are caught numerically. In color-managed projects (workingSpace != None) the capture applies AE's own display transform via a transient OCIO Display Transform adjustment layer, and the PNG comes back viewer-accurate and sRGB-tagged. Where that layer cannot be used — read-only mode, an AE without the OCIO effect, or a failed capture calibration — the result falls back to a pure-math ACES conversion or to raw values with an explicit colorWarning; check colorPipeline / colorWarning on the response.

Save the project. Pass path for Save As.

Serialize the entire project to JSON (folders, comps, layers, keyframes, effects, shapes, markers, time remap, solids, file refs). Write to outPath or return inline.

Rebuild the project from JSON (produced by ae_project_export_json). Supports clearFirst, dryRun (validate-only), skipValidation.

AE version, build, capabilities (saveFrameToPng, app.effects, Socket). Call at session start to know what APIs are available.

Discover available atomic operations for ae_do. Without args: all categories with their operation names. With a category: detailed params per operation. Only operations this server will actually execute are listed.

Execute an atomic operation by name (from ae_catalog). Arguments are validated against the operation's declared parameters before anything reaches AE. Have more than one operation to run? Wrap them in ONE batch.run instead of several ae_do calls — read/verify steps (comp.info, layer.info, render.frame) can ride in the same batch. Every call is automatically wrapped in ONE undo group: a single Ctrl+Z (or project.undo) reverts the entire call, and a batch.run counts as one call. Never call app.beginUndoGroup/endUndoGroup in eval.run code — the wrapper already did. The exception is undo/redo itself (project.undo, command.execute id 16/2035): those run outside the group and cannot ride inside a batch.run. Response includes ambient context (active comp, selected layers, project state) at zero extra round trips. Example: ae_do({ operation: 'keyframe.add', args: { comp: 'Main', layer: 1, property: ['Transform','Position'], time: 2, value: [960,540] } })

Ambient context: project state, active comp, selected layers, item list, AE.* helpers, ES3 rules, and the undo contract (every call = one auto undo group). Call at session start; then rely on ae_do response context.

The motion-design and studio knowledge this MCP carries: workflow rules, easing and timing, kinetic typography, PSD-to-animation recipes, the user's installed plugins / transitions / presets / MOGRTs and when to use them, client brand rules. Call with no arguments to list topics, topic to read one, query to search all of them. Read the relevant topic BEFORE planning a non-trivial animation or choosing an effect, so the plan uses what is actually installed.