Orchestrates ArcReel video projects through plan-driven workflow steps via MCP tools.
npx skills add https://github.com/ArcReel/ArcReel --skill video-workflow
Please help me install this Agent Skill. Skill: video-workflow Source: https://github.com/ArcReel/ArcReel/tree/main/skills/video-workflow Command: npx skills add https://github.com/ArcReel/ArcReel --skill video-workflow
Managing an AI video production project involves many ordered steps—creating the project, generating assets, storyboards, narration, and video clips—and it is easy to lose track of what to do next or accidentally re-run paid generation tasks. This Skill routes every request through the ArcReel workflow plan so each project advances correctly, with billing-sensitive actions confirmed before execution.
get_workflow_plan and executes only the action specified by next_action, keeping project state consistent with the ArcReel MCP server as the single source of truth.batch_id values and polls get_generation_batch until completion, with bounded retry limits for long-running tasks.Continue my ArcReel video project and tell me what the next step is, then run it after I confirm.
High-intent search queries and answers about installing and using this skill.
Ask to continue the project and the Skill calls get_workflow_plan to read next_action, then executes only that action with the plan's arguments. Actions requiring confirmation or billing are explained first and wait for your explicit approval.
Ask for a progress check and the Skill reports the plan's steps, problems, and next_action without modifying the project. It distinguishes workflow step status, task status, vendor submission checkpoints, and artifact status.
No, the remote MCP workflow does not compose or export final videos. When next_action is export, the Skill hands off to the WebUI or embedded host, which handles final rendering and export.
The Skill keeps the batch_id and polls get_generation_batch at each returned poll_after_seconds interval until done is true. If a plan returns wait_for_task, it re-polls up to max_poll_attempts, then reports the running task_ids and stops.
No, stale artifacts remain usable and are never regenerated or deleted automatically. Regeneration of current or stale artifacts happens only when you explicitly select those IDs, protecting paid generation history.
Do not use it for editing a single existing artifact in isolation; it is designed for end-to-end project orchestration. It also cannot repair project data directly when remote tools lack that capability.
punkfuncgamesAlign Unity in-game visuals with reference designs via MCP bridge workflows.
pawbytesAutomate short-form video production from brief to subtitled MP4.
LaiTszKinGenerate long-form 16:9 videos with storyboard images, narration, and Remotion rendering.
icyzhExecute generative AI workflows for images, video, and audio via ComfyUI APIs.