Photo mosaic maker and AI photo collage maker: turn your own photos into a mosaic or a collage.
- Passed: Handshake
- Passed: Tools listed
- Passed: Updated recently
Median latency 343 ms · Last checked
Explore
Evidence for your next connection.
Independent measurements of remote MCP servers. Every observation has a timestamp.
Read-only probesMCP server · AI / LLM tooling
Find video editing techniques by the result you want, with prompts, sound effects and templates.
1 failed · 7 not measured
Median of successful probes over 7 days
Share of countable probes over 30 days, including daily rollups. Auth challenges are excluded.
Tool definitions returned by this server
Independent monitoring
Responded
The server completed its last measured handshake. This is an observation, not a guarantee.
Last checked| Test area | Passed | Failed | Unmeasured |
|---|---|---|---|
ConnectionCan a client connect and discover tools? | 5 | 0 | 0 |
Reliability & speedDoes it respond consistently and quickly? | 1 | 0 | 3 |
Access & authenticationWhat is required to use this server? | 1 | 0 | 2 |
Tool qualityAre the tool definitions usable by an agent? | 3 | 1 | 0 |
Publisher identityCan the publisher be traced to its source? | 4 | 0 | 2 |
Unmeasured means there is not enough evidence to decide. How these checks work
Start using EditAtlas
Choose your client, copy the configuration, then enable the server in your client’s MCP settings.
Run this command in your terminal, then restart Claude Code.
Claude Code
claude mcp add --transport http editatlas https://edit.trysometh.ing/mcpCapabilities
Get one sound effect with its download links and checksum. github_download_url is the same file on GitHub, for sandboxes that cannot reach download_url. guide says where to put it on the frame and at what level, read from this sound's own waveform and loudness.
DiscoveredList the EditAtlas skills, or get one. A skill is a folder of instructions and scripts that runs a whole step on the user's machine, for example sound-design places library sounds on a finished video and checks sync and loudness, and transcribe makes a checked word-level transcript with Whisper for captions and cuts. Without name you get every skill, or only those for one library. With name you get its install steps for Claude Code, Claude apps and other agents, links pinned to the released version, and, with include_instructions, its SKILL.md text.
DiscoveredRead a versioned creative style and its compatibility constraints. User instructions and template hard constraints take precedence.
DiscoveredGet the full prompt for one technique and its parameters. has_tested_ffmpeg is true when the prompt's FFmpeg command has been run by EditAtlas, and false when it has no command or the command wasn't run. guide, when present, says what the technique needs first, such as a word-level transcript.
DiscoveredRead a versioned editing template with inputs, conditional beats, referenced techniques and acceptance checks. This is reference data, not an executed edit. guide, when present, says how to get an input such as a verified transcript.
DiscoveredResolve a template and an optional compatible style, with deduplicated techniques and public SFX. Compact by default; full includes technique prompts. Bounded to 40 dependencies and 96 KiB; sound queries remain suggestions.
DiscoveredGet how to run a whole edit, from inspecting files to delivery. Pass video_type for the structure, pacing, first techniques and the templates, music or stock media worth searching for that kind of video. If a job is two kinds at once, call it again for the second kind with include_workflow set to false. guide lists the next calls for that kind of video, including skills that do a whole step.
DiscoveredFind reusable creative constraints across typography, color, rhythm, motion and sound. Styles do not include assets or rendering services.
DiscoveredSearch background music tracks. Write query in English, such as "calm piano" or "upbeat corporate"; Vietnamese and other languages give poor matches, so translate first. genre narrows the search, for example "intro outro", and can be used without a query. duration is one of the length ranges 0-30, 30-120, 120-240 or 240-480 seconds; pick the one that holds the length the edit needs. Each result has a direct mp3_url, the author, duration_seconds and two flags from the source: has_youtube_content_id (a YouTube upload may get a Content ID claim; tell the user) and is_ai_generated. 20 results per page; raise page for more.
DiscoveredSearch stock photos on Pexels (default) or Pixabay. Write query in English, such as "coffee shop" or "city night"; translate other languages first. orientation is all, landscape, portrait or square; Pixabay photos take no square. Each result has width, height, a full-size image_url, a smaller preview_url, the author and credit_url. Tell the user who made what you use. Page with pages from the answer: Pixabay drops photos of the wrong shape, so a page can hold fewer than per_page.
DiscoveredSearch sound effects in English or Vietnamese by source and measured traits, such as sharp attack, long tail or wide stereo. Each result has a description, a preview and download links. Use get_sfx for the checksum and file details. For a hit under speech, pass duration tick or short. lufs is null when a sound is too short to measure. guide says how to choose among these results and what to call next.
DiscoveredSearch editing techniques by the result you want or by name. Returns about 10 candidates with short descriptions, for you to choose from. Words are matched all together first, then any of them.
DiscoveredFind composed editing blueprints by intent. Returns compact constraints; use get_template or get_template_bundle after choosing. Your agent performs all editing locally.
DiscoveredSearch stock video clips (B-roll) on Pexels (default) or Pixabay. Write query in English, such as "pouring coffee" or "drone city"; translate other languages first. orientation is all, landscape, portrait (for 9:16) or square; Pixabay videos cannot be filtered by orientation. Each result has width, height, duration_seconds, an mp4_url at the largest size, video_files at every size (pick the one that matches the edit), a thumbnail, the author and credit_url. Page with pages from the answer.
DiscoveredSend the user's own verdict on something EditAtlas gave you, after they have seen the result. Ask about each thing you used by name, in one short message (for example: "The film grain: right amount? The whoosh on scene changes: does it fit?"), and send one call per item the user answered about, with about this_item. If they answered only about the video as a whole, send that once, with about whole_edit, on the workflow you followed (or the main item); never copy one answer onto every item. Send only what they said; never rate on their behalf, and send nothing if they did not answer. For partly or failed, set issue and say what to fix: the user's words in note, what you saw in agent_note. Sending the same item again in this session replaces the earlier answer. Stores no footage, file names, studio links or personal details.
DiscoveredRecent measurements
| Checked | Result | Response | Tools |
|---|---|---|---|
| Connected | 274 ms | 15 | |
| Connected | 346 ms | 15 |