motion-reel-ingest
Skill1 starsChanged 3 months agoMITView on GitHub
Prepare mixed raw-video batches and scripts for motion-reel, classify and rename takes after map approval, convert HDR once, and manage a continuous Windows inbox watcher that starts Claude Code only for new ready batches. Use for incoming footage folders, batch intake, watcher setup, status and recovery.
In this repository24 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- ai-automation-contractsSkill
- brand-mascotSkill
- business-ai-auditSkill
- content-diarioSkill
- fable-5-protocolSkill
- hook-generatorSkill
- kieai-minimax-h3Skill
- lead-finderSkill
- mercadolibre-publisherSkill
- motion-reel-pipelineSkill
- motion-reelSkill
- pdf-markitdownSkill
- post-for-meSkill
- reel-pipelineSkill
- reels-referentesSkill
- replykaro-automation-communitySkill
- replykaro-automationSkill
- resource-pageSkill
- social-carousel-generatorSkill
- social-video-producerSkill
- soloduenos-leadsSkill
- video-animation-workflowSkill
- web-builderSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 12
Read BATCH.md when processing footage.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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
Motion Reel Ingest
This skill owns batch preparation and the AI-free inbox watcher. motion-reel owns video production; motion-reel-pipeline owns the full workflow order. A single edited video or already organized takes can go directly to production.
Intake
Read BATCH.md when processing footage. Use this skill's scripts/batch-takes.py. Scan/transcription, classification against the script, resource identification, map review, reversible renaming and shared SDR conversion belong here. Never rename before the user approves the actual map. Never delete discarded takes.
For a watched batch the watcher has already moved the whole folder from Crudos to En proceso/<batch-id>. Work there; preserve the ID. Runtime batch data is <batch>/.ingest/. For legacy work with existing <Documents>/motion-reel/.work/_lote-<name> state, reuse that state rather than retranscribing or renaming again.
Write one brief per topic under .ingest/briefs/: topic, absolute take paths by part, actual hook numbers, body/CTA order, script excerpt, confirmed resource URL, uncertain words and recording issues. Use SDR copies where available. Hand the briefs to motion-reel through the pipeline. Hook animation reuse belongs to motion-reel/reference/VARIANTS.md.
For watched jobs use <Videos>/Motion Reel/Entregas/.work/<unique-topic-slug> for production projects. The existing production packager then writes <Videos>/Motion Reel/Entregas/<unique-topic-slug>. Avoid collisions across batches; keep existing projects where they are. The source batch stays in En proceso through all review and publishing stages.
Watcher setup and operation
Installation includes setup and activation. Do not report the skill fully installed after only copying its folder. The repository installer runs scripts/install-watch.ps1 automatically after installing the package on native Windows in the default Claude skills directory. If installing by copying files, the installing agent must run that script itself. It discovers Claude Code, installs it through WinGet if missing, checks authentication, creates the Videos folders, registers/starts the task and verifies that it stays running. If Python is missing, the installing agent installs Python 3.10+ and reruns setup. Install all pipeline dependencies and check their requirements as part of onboarding. User browser authentication cannot be fabricated: mark setup pending login until it succeeds. Preserve existing tasks/jobs; do not interrupt active production during updates.
Read WATCHER.md for setup, readiness, commands and failure recovery. The script never calls a model to check folders. The Windows task keeps it alive; Claude Code runs only for a queued batch or an explicit user continuation. Polling is continuous every 10 seconds, with 60 seconds of unchanged files after LISTO is present. No daily AI automation is used.
The watcher starts motion-reel-pipeline in a separate Claude Code session with installed user skills. Preserve skill approval gates and configured tool permissions. Missing credentials, CLI, dependencies or MCP connections are setup failures; do not bypass permissions to proceed.
At every human gate, write <batch>/REVIEW.md with the actual review material's absolute paths, concise findings and exact question, then return. The watcher stores the final response and session ID. Show the material to the user when they open the session. Continue through watch.py resume with their actual reply, or open the saved session interactively while the watcher is idle for that job. Never run both on the same session at once.
Only after the user confirms the entire job is complete, use archive --confirm-complete to move the source batch to Archivados. Moving source files changes their absolute paths: reopening an archived production project requires restoring the batch to its original En proceso/<id> location first. Work projects and deliveries are retained. Failures and pending reviews remain in En proceso.
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.

