video-os-v2
AGENTS.md4 starsChanged 2 months agoMITView on GitHub
RoughCut Agent — artifact-driven video editing agent for turning source footage and a short brief into analysis, retrieval, rough/fine planning, compiled timelines, rendered outputs, and QA feedback loops. Required local toolchain: Project Loop Harness and the project-control-loop skill are temporarily paused for this repository. Do not invoke the skill or run pcl commands unless the user explicitly asks to re-enable or use them. The installed .project-loop/ state and pcl.yaml are retained unchanged so the harness can be re-enabled later. Never…
- Reads credentials
- Installs packages
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,439 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository23 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- analyze-footageSkill
- build-blueprintSkill
- compile-timelineSkill
- crop-photosSkill
- design-intentSkill
- evaluate-editSkill
- export-premiereSkill
- finish-business-shortSkill
- finish-creator-shortSkill
- finish-interviewSkill
- full-pipelineSkill
- import-premiereSkill
- re-editSkill
- render-videoSkill
- review-roughcutSkill
- select-clipsSkill
- select-photosSkill
- setup-environmentSkill
- short-sound-designSkill
- troubleshoot-errorSkill
- vertical-social-compositionSkill
- vertical-social-platform-deliverySkill
The file
We split this AGENTS.md into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Very specific: line 33
Treat canonical artifacts as truth: creative_brief.yaml, selects_candidates.yaml, edit_blueprint.yaml, timeline.json, render-report.json, …
- Distinctive: line 34
Keep Qwen3-VL, CLAP, Marlin, and other local-model paths fail-open when model cache or optional dependencies are absent.
- Very specific: line 35
Do not commit .env.local, source footage, rendered media, or generated project outputs unless explicitly requested.
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
AGENTS.md
Project
RoughCut Agent — artifact-driven video editing agent for turning source footage and a short brief into analysis, retrieval, rough/fine planning, compiled timelines, rendered outputs, and QA feedback loops.
Setup
npm install
cp .env.example .env.localRequired local toolchain:
- Node.js 22.x LTS (CI parity; see
.nvmrc/.node-version) - npm 10.x
Common Commands
| Task | Command |
| Demo | npm run demo |
| Test | npm test |
| Typecheck | npx tsc --noEmit |
| Build | npm run build |
| Verify | npm run verify |
| Full verify | npm run verify -- --full |
Hard Rules
- Preserve unrelated dirty worktree changes.
- Treat canonical artifacts as truth:
creative_brief.yaml,selects_candidates.yaml,edit_blueprint.yaml,timeline.json,render-report.json, QA reports, andfootage.db. - Keep Qwen3-VL, CLAP, Marlin, and other local-model paths fail-open when model cache or optional dependencies are absent.
- Do not commit
.env.local, source footage, rendered media, or generated project outputs unless explicitly requested. - For review-only or docs-only tasks, do not modify runtime code.
Where Agents Should Look
- North Star:
goal.md - Durable plan and handoff:
docs/project-memory/ - README and CLI entrypoints:
README.md - Architecture:
ARCHITECTURE.md - Planning route guidance:
docs/planning-routes.md - Source:
runtime/,scripts/,apps/,editor/ - Schemas:
schemas/ - Tests:
tests/,editor/tests/
<!-- project-loop-harness:start -->Project Loop Harness (temporarily paused)
Project Loop Harness and the project-control-loop skill are temporarily
paused for this repository. Do not invoke the skill or run pcl commands
unless the user explicitly asks to re-enable or use them.
The installed .project-loop/ state and pcl.yaml are retained unchanged so
the harness can be re-enabled later. Never hand-edit their generated or
database files while the harness is paused.
<!-- project-loop-harness:end -->Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.