No user-submitted related servers found.
Scored across 12 tools
Most tools have clearly distinct purposes: info readers target specific AE objects, render/save/export/import are unambiguous, and ae_catalog/ae_do form an obvious discovery/execution pair. The only real overlap is the trio ae_project_info, ae_version_info, and ae_context, all of which claim to be 'call at session start' and return partially overlapping project state, which could cause a redundant call.
The ae_ prefix is applied uniformly and names are readable, but conventions are mixed within that prefix: some are noun_style (ae_project_info, ae_comp_info, ae_layer_info), some are verb_noun (ae_render_frame, ae_save_project), and some are bare verbs or nouns (ae_do, ae_catalog, ae_guide). Predictable enough to navigate, but not a single consistent pattern.
Twelve tools is well-scoped for a Creative-After-Effects bridge: a small set of read/render/persist tools plus ae_catalog+ae_do as a deliberate escape hatch for the long tail of operations. Nothing feels padded, and the generic dispatch tool avoids an explosion of per-operation endpoints.
The surface covers the full lifecycle — inspect (project/version/comp/layer), mutate (ae_do), verify visually (ae_render_frame with contact sheets and analysis), persist (save, export/import JSON), and learn (ae_guide, ae_context). Minor gaps exist around explicit undo/redo and error recovery entry points, but ae_do's project.undo/command.execute coverage means agents are unlikely to hit a dead end.