edets_accountAccount
See who's connected and remaining seconds. The agent answers in plain language — not raw JSON alone.
The edets MCP server exposes video generation to any agent that speaks the protocol. Claude, Cursor, Grok Bot — they can generate clips and run an Ad Pack: five scenes plus a basic first cut.
https://edets.com/api/mcpURL → Allow → Connected{
"mcpServers": {
"edets": {
"url": "https://edets.com/api/mcp"
}
}
}Sign in on edets.com — don't paste a key. Manage sessions at edets.com/settings/mcp. Setup notes at edets.com/docs/mcp.
Restricted keys from edets.com/api/keys still work for clients without OAuth. Send Authorization: Bearer edets_live_…. Never the happy path. Ad Pack tools are OAuth-only and will not appear on a Bearer-key connection.
{
"mcpServers": {
"edets": {
"url": "https://edets.com/api/mcp",
"headers": {
"Authorization": "Bearer edets_live_…"
}
}
}
}edets_accountSee who's connected and remaining seconds. The agent answers in plain language — not raw JSON alone.
edets_generate_video_and_waitAttach a local still as a data URL (data:image/jpeg;base64,…) or send a public https URL. Describe motion, pick edets Max or edets Motion. Aspect includes 16:9. mute=true returns a silent clip. Cooks and returns when the file is ready.
agent: "Make a 5-second clip of this product photo. The model turns and looks at camera. Use edets Max at 720." edets: Cooking your clip… Succeeded in 34s. Download: https://edets.com/api/v1/generations/…/clip
edets_job_statusStill cooking or ready. Failed jobs return seconds to the pool.
generate_videoAlias for one release. Queues a clip and returns a generation ID. Prefer Create clip and wait.
create_campaignGenerate multiple shots, then compose a master and export platform packs (ig_9x16, tt_9x16, fb_4x5).
compose_campaignAfter clips succeed, stitch the campaign master and platform packs. Campaigns do not auto-compose.
Agent reads your product sheet, generates 6 clips in different styles, and drops them in your Slack for review.
Agent sees a support ticket, generates a personalized walkthrough video, and drafts the reply with the clip attached.
Agent takes your blog post, extracts 5 key points, generates one video per point, and schedules them across platforms.
Agent generates the same scene with 3 different motions, downloads all three, and reports which one it predicts will perform best.
Open source. MIT licensed. PRs welcome.