samuelgursky/davinci-resolve-mcp

v2.24.0

V2 trust-by-default media analysis: host-chat vision protocol, control panel review surface, correction tools, and analysis memory layer.

Added

• Host-chat vision protocol (V2) — analyze_* actions now use vision.provider="host_chat_paths" by default. The analyze response is a deferred payload with absolute frame_paths, a per-shot shot_table, and a JSON schema; the host chat reads each frame as a local image, produces JSON per the schema, and calls media_analysis(action="commit_vision", params={clip_id, visual, vision_token}) to finalize. commit_vision merges the visual report, rebuilds Media Pool clip markers, publishes metadata to Resolve, and preserves human corrections via corrections.json. Skipping the commit leaves the run in pending_host_vision_analysis — surfaced explicitly rather than silently downgraded. Legacy chat_context / mcp_sampling providers resolve to the same host-chat path.
• Trust-by-default analysis defaults — analyze_media defaults to include_transcription=true, persist=true, publish_metadata=true, and timed_markers="ask". Agents that need a technical-only or read-only run must explicitly pass the opt-out flags. The analyze_media prompt and SKILL.md encode the anti-regression rule.
• Control panel: Review surface (Phase B) — Bin grid with thumbnails, clip detail with shot strip, shot detail with grouped V2 fields + frame grid, inline correction editors per subjective field, transcript correction + regeneration, Open-in-Resolve bridge, and 2-second chat ↔ panel state polling. Backed by new endpoints: /api/clips, /api/clips/<id>, /api/clips/<id>/shots, /api/clips/<id>/shots/<index>, /api/clips/<id>/frames/<n>, /api/clips/<id>/transcript, /api/clips/<id>/corrections, /api/clips/combined, /api/clips/export, /api/panel_state, /api/update/status, /api/update/apply, /api/resolve/open_clip.
• Control panel MCP actions — resolve_control.open_control_panel, control_panel_status, close_control_panel manage the panel subprocess via a pidfile. save_state / restore_state snapshot and restore Resolve playhead + selection state. get_panel_state, set_panel_state, session_start_context share panel focus between chat and the UI through panel_state.json.
• Correction tools — New media_analysis actions: update_shot_field, update_clip_field, get_field_history, revert_field, list_corrections. Writes land in {clip_dir}/corrections.json with append-only changelog + provenance (mirrors V2 DB schema). commit_vision merges corrections on top of the fresh visual report so human edits survive re-analysis.
• Media Pool item open-in-viewer — media_pool_item.open_in_viewer selects a clip on the Media page and loads it into the source viewer, optionally setting mark in/out and bringing Resolve to the foreground via OS-level window activation.
• Source-trust prompt grading — source_trust parameter (auto / filename / low / medium / high) on analyze actions tunes the vision prompt to hedge identity/intent/value for archival or thin-evidence clips.
• Analysis memory layer — src/utils/analysis_memory.py introduces per-project memory + heartbeat + bin summary + soul scaffolding under the analysis root. regenerate_bin_summary_from_manifest aggregates per-clip fields into a bin briefing. Auto-initialized on analyze.
• Control-panel polish — Diagnostics + Overview restyled with a status-pill design system, navbar dropdowns fixed so top-level buttons no longer navigate on their own, Preferences consolidated, summary-style enum renamed to full / concise / creative / technical (with backwards compat), navbar version badge + update modal, source-trust dropdown wired through.

Fixed

• commit_vision auto-publish reflects per-row status correctly — no more silent-lie pending status.
• Compact analyze responses by default; pass verbose: true for the full manifest.
• resolve_output_root skips slug append when the configured base already terminates in the slug.
• Frame sampler reserves per-shot budget so shot starts aren't starved by flash candidates.
• Machine markers are no longer written to Resolve (V2 architecture).
• media_analysis now augments PATH with standard tool dirs (/opt/homebrew/bin, /usr/local/bin, etc.) so ffprobe/ffmpeg resolve when the MCP server is launched by a GUI app (Claude.app, Dock/Spotlight) that inherits launchd's bare PATH.

Documentation

• AGENTS.md adds the "Media Analysis Defaults Are Mandatory" section.
• docs/SKILL.md rewrites the analyze_media prompt guidance for the host_chat_paths protocol and adds the anti-regression rule.
• docs/guides/media-analysis-guide.md covers the deferred vision payload and commit step.
• New design docs under docs/design/: v2-control-panel-design.md, v2-db-schema.sql, v2-implementation-gameplan.md, v2-shot-schema-spec.md, control-panel-polish-gameplan.md.

Validation

• Static import checks (tests/test_import.py) — OK.
• API parity audit (scripts/audit_api_parity.py) — PASS.
• Focused unit tests (138 tests across test_extract_source_frame_ranges, test_marker_params, test_v232_helpers, test_v233_helpers, test_append_clip_infos_result_handling, test_media_analysis) — OK.
• npm CLI smoke (node bin/davinci-resolve-mcp.mjs --help/--version) — OK.
• npm pack --dry-run — OK (97 files, package version 2.24.0).
• git diff --check — clean.
• No source media was modified. Resolve scripting behavior is additive (new actions; existing actions unchanged); live Resolve validation covered the control panel + open_in_viewer + commit_vision auto-publish + corrections-merge paths during the V2 push sessions.