Agent skill

Resolve Edit

by samuelgursky in samuelgursky/davinci-resolve-mcp

Editing, cutting, trimming, pacing, and timeline restructuring in the DaVinci Resolve MCP.

MITAuto-check passedAgent Workflows

Install Resolve Edit

skills CLI
$ npx skills add samuelgursky/davinci-resolve-mcp --skill resolve-edit -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install samuelgursky/davinci-resolve-mcp resolve-edit --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/samuelgursky/davinci-resolve-mcp.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/resolve-edit .claude/skills/resolve-edit && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
resolve-edit
GitHub stars
3.4k
Token cost
~2.4k tokens
SKILL.md length
1,229 words
Files
1
Skills in repo
14
Repo updated
First seen
Licence
MIT

At a glance

Editing, cutting, trimming, pacing, and timeline restructuring in the DaVinci Resolve MCP.

  • Tasks that involve MCP servers
  • SKILL.md covers Two servers, What this build cannot do…, Live edit-kernel essentials and Show the gaps before cutting…, plus 3 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Resolve Edit is an agent skill from samuelgursky/davinci-resolve-mcp. Editing, cutting, trimming, pacing, and timeline restructuring in the DaVinci Resolve MCP. Apply when duplicating/moving clips, copying ranges, building variants, tightening or restructuring a cut, editing selects, or generating an editorial changelist/turnover — live in a running Resolve OR offline against .drt timeline files. Routes to the live edit tools, the offline editorial tools, and the project's editorial craft guidance.

Its SKILL.md is about 2.4k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Agent Workflows, covering MCP servers. It works with Model Context Protocol. The repository describes itself as: MCP server integration for DaVinci Resolve Studio. The licence is MIT.

When your agent uses it

  • Tasks that involve MCP servers

Example prompts

  • “/resolve-edit”

What it can do on your machine

Read from SKILL.md and the folder at commit a733574. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Resolve Edit loads about 2.4k tokens when it runs. Until then it costs about 112 tokens; SKILL.md has 1,229 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~112
When it runs · the whole SKILL.md, loaded when a task matches
~2.4k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from samuelgursky/davinci-resolve-mcp at commit a733574, republished under its MIT licence (© samuelgursky). 1,229 words, ~2,422 tokens.

Download SKILL.mdSave it as .claude/skills/resolve-edit/SKILL.md (or your agent's skills folder).
name
resolve-edit
description
Editing, cutting, trimming, pacing, and timeline restructuring in the DaVinci Resolve MCP. Apply when duplicating/moving clips, copying ranges, building variants, tightening or restructuring a cut, editing selects, or generating an editorial changelist/turnover — live in a running Resolve OR offline against .drt timeline files. Routes to the live edit tools, the offline editorial tools, and the project's editorial craft guidance.

Resolve Timeline Edit

Bridges editorial craft to this repo's tools. Open the right manual; don't re-derive it here.

  • Craft / story — docs/guides/editorial-decision-guide.md. The global editor / assistant-editor skills add editorial philosophy and cutting-room practice; use them for why to cut, not tool mechanics.
  • Live tool mechanics — docs/kernels/timeline-edit-kernel.md (the timeline edit-kernel boundary: duplicate/copy/move, range ops, item-state copy).
  • Offline timeline / changelist — resolve-advanced/README.md → drt (timeline file authoring) and editorial (interchange + turnover).

Two servers

JobServerTools
Restructure a running timelinedavinci-resolve (Python, live)timeline (edit kernel), timeline_item, edit_engine, timeline_markers
Author/diff a .drt file, or parse/compare editorial interchange with no Resolve opendavinci-resolve-advanced (Node)drt, editorial

What this build cannot do (check before you offer it)

The scripting API changes per patch release, so "Resolve 21" is not a usable label. Read resolve_control get_version → build.unavailable_on_this_build before offering a gated surface; check_version_support asks about one named symbol. Gated in this domain:

SurfaceNeedsIf absent
Timeline.GetSelectedClips21.0.4No selection readback. Identify clips by track/index or id instead — this repo's selection helper duck-probes three names, so it degrades to "no selection" rather than erroring
TimelineItem.SetName20.2Clip names are read-only from a script

An empty unavailable_on_this_build means nothing recorded is missing, not that everything exists — most of the API has never been version-bisected. A symbol with no gate returns unknown, which means probe it. Probe with name in dir(obj), never bare hasattr: on a Resolve object hasattr returns True for every name, real or invented, so it can only say yes.

Distinguish gated from absent. A gated surface arrives with a newer build; an absent one never does, and upgrading will not help. Clip speed / retime is absent on every build — see the edit-kernel essentials below before you offer set_retime for it.

Live edit-kernel essentials

  • Duplicate/relocate: duplicate_clips (modes same_time/offset/ at_playhead/track_above/after_source/next_gap), copy_clips (alias), move_clips (duplicate-then-delete). include_linked=True carries linked audio.
  • Ranges: copy_range, duplicate_range, overwrite_range, lift_range. No public razor/split — partial overlaps in lift_range are blocked unless allow_partial_item_delete=True (whole-item delete, not a trim).
  • Item state copy: copy_properties (transform/crop/composite/audio/retime/ markers/flags/grades/takes/keyframes …); scope with a group list.
  • Reading the user's selection is best-effort and version-dependent. The selection helper duck-probes three method names; the documented one, Timeline.GetSelectedClips, is 21.0.4+ (issue #131), so on older builds selection resolves by luck or not at all. Never build a destructive operation on "the selected clips" without reading back what you actually got — an empty or partial selection is indistinguishable from a small one.
  • There is no clip-speed API at any version. set_retime sets retime quality only (RetimeProcess, MotionEstimation) and returns True for doing so, which reads as if the retime succeeded. Setting a % speed, reversing a clip, and speed ramps are all unreachable — SetProperty('Speed'|'PlaybackSpeed'| 'RetimeSpeed'|'ClipSpeed') returns False and the matching GetProperty returns None on every build measured. Say so and route the user to the UI; do not offer set_retime as if it answered the question (issue #132).
  • Routing to the UI is where issue #132 actually went wrong — and it was not a tool lie. Having been told to hand off, the assistant improvised the handoff: it sent the user hunting for a dropdown "in the lower left of the clip" and never mentioned the keyframe tray. That direction came from nowhere in this repo. This repo verifies API behavior, not UI geography, and has no mechanism to version-guard a UI claim — Resolve's controls move between builds, pages and layouts, and every ledger entry here is stamped with the build it was measured on precisely because unstamped claims rot. So: never improvise UI geography. Say what the user needs to do ("set the clip's speed / build the speed ramp in Resolve's own retime controls") and say plainly that you cannot see their screen. Do not invent where a control sits, what it looks like, which corner of a clip to click, or a menu path or keyboard shortcut recalled from general knowledge — an invented location costs the user more than "I don't know where it is on your build" ever does, and it reads as authoritative because everything else you told them was measured. The line is improvised vs. written down here. A UI pointer that already exists in these skills (e.g. the playback-frame-rate path in resolve-rough-cut) was authored and reviewed deliberately — use it verbatim. If no pointer exists, that is not an invitation to supply one: name the operation and point at Blackmagic's manual for their build. The same rule covers every other "do it in the UI" handoff in these skills.
  • edit_engine drives higher-level selects/tighten/swap flows (plan → confirm → execute); tighten variants can carry audio via keep_ranges mirror / include_audio. For a full dead-air pass over one long single-take recording — transcribe → plan → review → execute, with the source/record coordinate trap spelled out — use the resolve-tighten-recording skill.
Show full SKILL.md (452 more words)Show less

Show the gaps before cutting them (edit_engine)

  • plan_dead_space_markers(timeline_name?, tightness?) — the review gate. Same calibrated detection as plan_silence_ripple, but it proposes markers instead of assembling a variant, so a human sees every gap before agreeing to lose it. Use this whenever the ask is "mark the dead space so I can review and approve" — that is a request for this verb, not for a tighten. Red = confident, Yellow = the gate only just cleared its separation floor. Nothing is written; pair the marker specs with timeline_markers.
  • tightness on the dead-space planner: generous (default) | balanced | tight. Default is deliberately the loosest. A first assembly is supposed to be long — trimming a generous cut is fast and visible, while recovering material the machine already discarded is slow and invisible. Move to tight only when the editor asks for it.
  • Guard bands are floored regardless of tightness: a removed gap always leaves frames before it (the outgoing word's decay) and after it (the incoming word's attack). tight removes more gaps, never more speech.

Word-level editing and spoken search (edit_engine)

Silence-driven tightening cannot remove an audible "um" mid-phrase or a stammered restart. These work on the words instead and are complementary to plan_silence_ripple, not a replacement.

  • plan_transcript_tighten(clip_ref) — fillers, false starts and long pauses, at word boundaries. Emits the same keep_ranges shape the variant assembler already takes, and every removal states its reason, so the plan can be argued with rather than only accepted. False starts are a flagged heuristic: tight repeats only, so "no, no, no" as emphasis survives.
  • search_spoken_content(query, mode) — phrase / all_words / regex across every transcribed clip, returning timestamped hits with context plus selects (in/out with handles). A lexical axis; find_similar is the semantic-visual one. Use both — a shoot should be searchable either way. Deterministic order (clip name, then time), so an identical search gives an identical selects list.
  • Needs transcript_words: run transcription, then strata backfill. Both actions say so rather than returning an empty plan.

Offline editorial (editorial actions)

  • parse_interchange — EDL / OTIO / XMEML (AAF = an honest refuse, not a fake).
  • turnover_changelist — moved / retimed / trimmed / replaced / new / gone between two cuts PLUS the junction diff (transitions added / dropped / changed, fade in/out, span deltas — carrier lines and BL fade legs never read as sources), with timing silent-lie guards (it flags what it cannot verify).
  • conform_manifest, marker_roundtrip.

Use these to answer "what changed between v3 and v4" or to hand a conform an accurate change list without opening either timeline in Resolve. For carrying a conform across a re-edit, see the resolve-conform skill.

Source-media safety (AGENTS.md)

Edit operations reference existing Media Pool items — they never transcode, render, proxy, or create derivatives of source media. Keep it that way. Treat generated probe reports as local scratch artifacts, not committed files.

© samuelgursky, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .agents/skills/resolve-edit of samuelgursky/davinci-resolve-mcp.

Open the folder on GitHubat commit a733574

Compare with similar skills

Resolve Edit next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Resolve Edit compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Resolve Edit this skillsamuelgursky/davinci-resolve-mcp3.4k—~2.4kAutomated safety check: PassMIT
MCP Server Builderanthropics/skills180k63 repos~2.3kAutomated safety check: PassApache-2.0
MCP Server BuildershareAI-lab/learn-claude-code78k5 repos~1.2kAutomated safety check: PassMIT
MCP Integration for Pluginsanthropics/claude-plugins-official38k11 repos~3.1kAutomated safety check: PassApache-2.0
Crush Configurationcharmbracelet/crush29k—~3.7kAutomated safety check: PassCustom licence
Context Mode Output Sandboxmksglu/context-mode26k—~4.1kAutomated safety check: PassCustom licence

Similar skills

  • MCP Server Builder

    anthropics/skills

    Official

    Guides the design and implementation of Model Context Protocol servers in TypeScript or Python, from tool naming and error messages to evaluation.

    180k GitHub starsUsed in 63 repos~2.3k tokens
    Agent WorkflowsAuto-check passed
  • MCP Server Builder

    shareAI-lab/learn-claude-code

    Walks through building MCP servers in Python or TypeScript that expose tools, resources and prompts to Claude, with templates, registration and testing.

    78k GitHub starsUsed in 5 repos~1.2k tokens
    Agent WorkflowsAuto-check passed
  • MCP Integration for Plugins

    anthropics/claude-plugins-official

    Official

    Explains how to bundle Model Context Protocol servers in a Claude Code plugin, covering config files, stdio, SSE, HTTP and WebSocket server types, and authentication.

    38k GitHub starsUsed in 11 repos~3.1k tokens
    Agent WorkflowsAuto-check passed
  • Crush Configuration

    charmbracelet/crush

    Explains how to configure the Crush coding agent with crushrc or crush.json, covering providers, models, LSPs, MCP servers, hooks, permissions and config precedence.

    29k GitHub stars~3.7k tokensUpdated today
    Agent WorkflowsAuto-check passed
  • Context Mode Output Sandbox

    mksglu/context-mode

    Routes large command, file, API and browser output through context-mode tools so only the needed result enters the agent's context, instead of dumping it via Bash.

    26k GitHub stars~4.1k tokensUpdated today
    Agent WorkflowsAuto-check passed
  • Migrates the compatible subset of settings and global file-based MCP servers from the Warp desktop app into Warp Agent CLI without exposing credentials or state.

    65k GitHub starsUsed in 1 repo~2.1k tokens
    Agent WorkflowsAuto-check passed

More from samuelgursky/davinci-resolve-mcp

All 14 skills in this repo
  • Resolve MCP

    samuelgursky/davinci-resolve-mcp

    Orientation and index for DaVinci Resolve MCP work — grading, editing, conforming, delivery, media analysis, and .drp/.drt/.drx file work, live in a running Resolve or offline with none open.

    3.4k GitHub stars~1.5k tokensUpdated today
    Auto-check passed
  • Resolve Rough Cut

    samuelgursky/davinci-resolve-mcp

    Assembling a short-form social rough cut from raw behind-the-scenes or vlog footage in the DaVinci Resolve MCP.

    3.4k GitHub stars~2.9k tokensUpdated today
    Auto-check passed
  • House Style

    samuelgursky/davinci-resolve-mcp

    The editorial and finishing preferences this project's work is judged against — cut rhythm, shot selection, delivery conventions, and the corrections that have already been given.

    3.4k GitHub stars~736 tokensUpdated today
    Auto-check passed
  • Resolve Audio

    samuelgursky/davinci-resolve-mcp

    Audio and Fairlight work in the DaVinci Resolve MCP. An agent skill from samuelgursky/davinci-resolve-mcp.

    3.4k GitHub stars~1.3k tokensUpdated today
    Auto-check passed
  • Resolve Color

    samuelgursky/davinci-resolve-mcp

    Color grading and look work in the DaVinci Resolve MCP. An agent skill from samuelgursky/davinci-resolve-mcp.

    3.4k GitHub stars~1.9k tokensUpdated today
    Auto-check passed
  • Resolve Conform

    samuelgursky/davinci-resolve-mcp

    Conforming, relinking, and finishing prep in the DaVinci Resolve MCP.

    3.4k GitHub stars~2.1k tokensUpdated today
    Auto-check passed

Categories

Questions about Resolve Edit

What does Resolve Edit do?

Editing, cutting, trimming, pacing, and timeline restructuring in the DaVinci Resolve MCP. Resolve Edit is an agent skill from samuelgursky/davinci-resolve-mcp. Editing, cutting, trimming, pacing, and timeline restructuring in the DaVinci Resolve MCP.

When should I use Resolve Edit?

Resolve Edit fits situations like: tasks that involve MCP servers.

How do I install Resolve Edit in Claude Code?

Run `npx skills add samuelgursky/davinci-resolve-mcp --skill resolve-edit -a claude-code`. Or copy the skill folder (.agents/skills/resolve-edit in samuelgursky/davinci-resolve-mcp) into .claude/skills/resolve-edit in your project. Claude Code loads it when a task matches its description.

How do I install Resolve Edit in Codex?

Run `npx skills add samuelgursky/davinci-resolve-mcp --skill resolve-edit -a codex`. Or copy the skill folder (.agents/skills/resolve-edit in samuelgursky/davinci-resolve-mcp) into .agents/skills/resolve-edit in your project. Codex loads it when a task matches its description.

Can I use Resolve Edit in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add samuelgursky/davinci-resolve-mcp --skill resolve-edit -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/resolve-edit, .gemini/skills/resolve-edit, .github/skills/resolve-edit and .opencode/skills/resolve-edit in your project.

What does Resolve Edit need to run?

SKILL.md names no scripts, command-line tools or credentials: Resolve Edit is instructions for the agent only.

Does Resolve Edit access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Resolve Edit safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Resolve Edit use?

Resolve Edit is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Resolve Edit use?

About 2.4k tokens (SKILL.md is roughly 9.7k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Resolve Edit?

Skills that share tags, products or a category with Resolve Edit: MCP Server Builder (anthropics/skills, 180k stars), MCP Server Builder (shareAI-lab/learn-claude-code, 78k stars), MCP Integration for Plugins (anthropics/claude-plugins-official, 38k stars) and Crush Configuration (charmbracelet/crush, 29k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Resolve Edit?

samuelgursky (a GitHub user) maintains it in samuelgursky/davinci-resolve-mcp, which has 3,426 GitHub stars. The repository holds 14 skills in this directory. The repository was last updated on October 9, 2026.

Source: samuelgursky/davinci-resolve-mcp on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.