Writes a SHA-256 hash of the sort log into a Solana devnet memo transaction, and verifies a seal by recomputing the log hash and checking it against the chain. Only the hash goes on-chain — never file names or contents.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Cullyard scan my Bali folder and sort photos into Picks, Maybe, and Rejects"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Tell your agent how to sort. Watch the folders move.
Cullyard is a local MCP server with a desk in the browser. Claude Code, Antigravity or any other MCP client reads every file in a folder, sorts it into lanes (Picks, Maybe, Rejects, or lanes you name), and trims or resizes copies when you ask. The desk at http://127.0.0.1:4870 shows each move as it happens, and you can drag cards back if you disagree. When a sort is done, the agent can write a SHA-256 of the sort log into a Solana devnet memo, so anyone holding the log can check later that it was not edited.
Kind | Examples | What the agent sees |
image | jpg, png, webp, heic, avif, gif, tiff | width, height, format |
raw | cr2, cr3, nef, arw, dng, raf | width, height when the decoder can read it |
video | mp4, mov, mkv, webm, avi, mts | duration, codec, frame size |
audio | mp3, flac, wav, m4a, ogg, opus | duration, title, artist |
page count | ||
document | docx, xlsx, csv, pptx, md, epub | kind and size |
archive, design, code, other | zip, psd, blend, py, anything else | kind and size |
Unknown extensions are still listed as other. The desk never refuses a file because of its extension.
Related MCP server: FilePilot AI
You need Node.js 20 or newer. FFmpeg ships inside the package, so trimming works on a clean Windows machine.
Claude Code:
claude mcp add cullyard -- npx -y github:bryankwandou/cullyard "D:\Shoots\Bali"Antigravity, Cursor, Claude Desktop or any client that reads an mcpServers block:
{
"mcpServers": {
"cullyard": {
"command": "npx",
"args": ["-y", "github:bryankwandou/cullyard", "D:\\Shoots\\Bali"]
}
}
}Replace the folder with your own. The desk opens on 127.0.0.1:4870 (set CULLYARD_PORT to change it). Pass --no-desk if you only want the MCP tools.
To seal runs on devnet, set CULLYARD_KEYPAIR to the path of a devnet keypair JSON file. Without it, sorting and editing work the same and seal_run returns a clear error.
Tool | What it does |
| Lists every file with kind, size, pixel size, duration, tags or page count |
| Metadata for one file |
| Current lanes and which file sits where |
| Adds a lane, such as "Hero shots" or "Client docs" |
| Moves a file into a lane folder, with an optional reason shown on the desk; never overwrites |
| Measures focus, exposure and bursts for every photo, ranks them against each other |
| One numbered grid of up to 16 photos, returned as an image the agent can look at |
| One photo as an image (up to 1024 px), plus its measurements |
| Writes a resized WebP, JPG, PNG or AVIF copy |
| Cuts a time range from video or audio without re-encoding |
| SHA-256 of the sort log, computed locally |
| Writes that hash into a Solana devnet memo |
| Recomputes the hash and checks it against the chain |
| The file kinds Cullyard recognises |
Rules the server enforces: paths outside the workspace are refused, moves never overwrite, edits always write a new file, and the original is never touched.
Metadata alone cannot tell a sharp frame from a blurred one, so Cullyard gives the agent two kinds of evidence.
Pictures. contact_sheet and preview_image return real images over MCP, so a vision model such as Claude looks at the photos instead of guessing from file names. The sheet numbers each frame, and the legend maps numbers back to files.
Measurements. assess_photos measures each photo and compares it with the rest of the set:
Focus: variance of the Laplacian on a contrast-stretched greyscale copy, measured per tile on a 4×4 grid. The sharpest tile counts, so a sharp subject on a soft background still reads as sharp, and a dark frame is not mistaken for a soft one.
Exposure: mean brightness and the share of clipped shadows and highlights.
Bursts: a 64-bit difference hash plus a 3×3 colour print. Two frames are one burst only when both match, so the same framing at golden hour and at blue hour stays separate. The sharpest frame of each burst is kept; the others are flagged duplicate.
Flags: soft (under 40% of the set's median focus), dark, bright, duplicate. Rank puts clean frames first, sharpest first.
The results are saved in .cullyard/quality.json, and the desk shows them on each card, so you see the same evidence the agent used. The numbers are signals, not verdicts: a shallow-focus portrait or a night shot can be a keeper with "bad" numbers, which is why the agent also looks at the picture.
scripts/demo-agent.mjs drives the real server over stdio, the same way an agent does. It scans twelve sample files, creates a "Client docs" lane, measures the seven photos, looks at them on a contact sheet, sorts photos from what it measured (with a reason on every move), files the other four, resizes the top pick to 1200 px WebP, trims three seconds from a clip, seals the log on devnet and verifies the seal. The full output is in docs/agent-transcript.txt, and the recorded desk is in web/media/walkthrough.mp4.
npm install
node scripts/make-samples.mjs demo-workspace
CULLYARD_KEYPAIR=path/to/devnet-keypair.json node scripts/demo-agent.mjs demo-workspaceThe seal from the recorded run is in web/proof.json and can be opened on Solana Explorer.
Every move appends one JSON line to .cullyard/log.jsonl inside the workspace. seal_run hashes the log with SHA-256 and sends one transaction to the Memo program (Memo1UhkJRfHyvLMcVucJwxXeuD728EqVDDwQDxFMNo) with the text:
cullyard:v1 log=<sha256> entries=<count>verify_seal fetches that transaction, hashes the log lines written before the seal, and compares the two. File names and file contents never go on-chain; only the hash does.
npm testEighteen tests cover file kinds, scanning, lanes, path traversal, no-overwrite moves, resizing, video probing, the desk's upload and cross-origin guard, the seal digest, photo measurement (blur, a sharp subject on a soft background, dark frames, bursts, colour-separated framings), the contact sheet, and a full MCP stdio round trip that includes the image tools.
bin/ command-line entry (MCP over stdio, plus the desk)
src/ scan, lanes, edit, look (photo measurement), seal, MCP server, desk server
desk/ the browser desk (plain HTML, CSS, JS; English and Indonesian)
web/ landing page, screenshots, walkthrough video, proof.json
scripts/ sample generator and the demo agent
test/ node:test suites
docs/ report and agent transcript
pitch/ pitch deck
video/ marketing videoThe photo thresholds were tuned on generated test images, not yet on a real shoot. Treat flags as hints and check the contact sheet.
Image tools need an MCP client that shows image results to the model. Claude Code does; other clients were not tested.
There is no timeline editor. Trim and resize are single steps.
Seals use devnet only.
MIT
Measure sharpness (per region, so a sharp subject on a blurred background still scores well), exposure and clipping for each photo, find near-duplicate bursts and mark the sharpest frame of each, and rank the set. Flags are relative to the set: soft, dark, bright, duplicate. Treat them as evidence, then look with contact_sheet or preview_image before rejecting a photo.
Inputs: lane, files.
Return up to 16 photos as one numbered grid image you can look at, with a legend mapping numbers to files. Use it to compare a set: composition, faces, closed eyes, missed focus.
Inputs: lane, files, offset.
Create a custom lane, for example "Hero shots" or "Final cut".
Inputs: name (required).
Return kind and metadata for one file in the workspace.
Inputs: file (required).
List the file kinds Cullyard recognises.
Inputs: no named parameters.
Return the lanes and which file sits in which lane.
Inputs: no named parameters.
Show the current SHA-256 and entry count of the sorting log, without sending anything.
Inputs: no named parameters.
Move a file into a lane folder. Never overwrites. Give a short reason; the desk shows it to the person watching.
Inputs: file (required), lane (required), reason.
Return one photo as an image you can look at (up to 1024 px), with its measured quality. Use it to check focus on the eyes, expression or detail before deciding.
Inputs: file (required), size.
Write a resized copy of an image next to the original.
Inputs: file (required), width, format, height.
List every file in the workspace with its kind, size and metadata.
Inputs: no named parameters.
Write the SHA-256 of the sorting log into a Solana devnet memo, as proof of what was sorted and when.
Inputs: no named parameters.
Cut a time range out of a video or audio file without re-encoding.
Inputs: file (required), endSec (required), startSec (required).
Recompute the log hash at seal time and compare it with the memo on Solana devnet.
Inputs: signature (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 14 tools
Most tools target clearly distinct resources and actions: sealing (seal_run/verify_seal/log_digest), workspace inspection (scan_workspace/describe_file/list_kinds), lanes (list/create/move), and media editing (resize/trim) are well separated. The photo-inspection trio (assess_photos, contact_sheet, preview_image) has some conceptual overlap since all return quality information, though the descriptions distinguish set-level measurement, grid comparison, and single-image inspection.
Nearly all names follow a consistent snake_case verb_noun pattern (seal_run, list_kinds, scan_workspace, move_to_lane, assess_photos, resize_image). A couple deviate to compound nouns (contact_sheet, log_digest), which is a minor inconsistency but still readable and predictable.
14 tools is well within the sweet spot for this domain, covering sealing, inspection, lane management, photo assessment, and light media editing. Each tool has a clear and non-redundant role.
The surface covers the core lifecycle: scan/describe files, create/list lanes, move files, assess and preview photos, resize/trim media, and seal/verify logs. Minor gaps exist—no delete or remove-from-lane operation, no lane update/delete, and no batch move—but most workflows can be completed.
Solana MCP for wallets, trades, markets, PnL, transfers, onchain data, signable swaps and API tools.
Remote MCP for C2PA intake verifier MCP, structured receipts, audit logs, and reviewer-ready evidenc
Solana-native MCP gateway for SAP, DeFi tools, SNS identity, and x402 payments.
Provenance and metadata tools for AI agents. Paid tools use x402 (USDC) at /mcp/<tool>.
