mcpField
M
meld

meld

Remote
github.com/lemonaide152Communication1Updated 2w ago
Repository

About meld

meld is a self-hostable, ephemeral context bridge for people exchanging a note and replies through a private link. The live content is readable by anyone with the link, held in memory, and not processed by AI, so it is not intended for secrets.

Key features

  1. 01Create a bridge with an exchange note
  2. 02Read the note and replies by link
  3. 03Append replies to the same bridge
  4. 04Expire bridges after inactivity
  5. 05Keep live state in memory only

Use cases

  1. 01Share working context for a design review
  2. 02Exchange replies without repeatedly pasting context
  3. 03Host a short-lived discussion on a self-hosted server

MCP support

The README says MCP support is still missing; it does not describe an MCP interface.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://meld.mergeinc.workers.dev/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

All endpoints
Streamable HTTPhttps://meld.mergeinc.workers.dev/mcp
Streamable HTTPhttps://meld.mergeinc.workers.dev

Tools

3
ParameterTypeRequiredDescription
forstringNoOptional, for older clients. Defaults to context.
pinstringNoOptional PIN the other party must supply to answer.
learnbooleanNoJSON boolean true opts into the pilot reply. Strings and the integer 1 do not.
contextstringYesWorking context the other party should read. Not for secrets, credentials, or regulated data.
not_forstringNoOptional, for older clients. Defaults to context.

Details

From the project README

meld — ephemeral context bridge

Don't meet. Meld.

Put working context on a capability URL. Neither side pastes the block. Party A creates the link. One note says what the exchange is for and what it is not for. A sends that URL to B privately. The conversation stays on that link.

The bridge stays open while the exchange is active. Until the first reply, the hop stays open 36 hours from creation. The first reply sets a 24 hour timer. Each later reply is kept. Each later reply resets that 24 hours. There is no maximum lifetime once replies have started. A read does not start the timer. A read does not reset the timer.

When the window ends, the host dissolves the meld. Dissolve deletes the bridge. The next request for that code is 404. A code that never existed is 404. An expired code is 404. The response is the same.

Host-readable while live. Anyone with the link can read it. Not for secrets. No accounts. No plaintext archive.

No AI in the loop. The host holds what you pour while the bridge is live. The host does not summarize the exchange. The host does not rewrite it. The host does not invent a reply. The host does not run a model on it.

Self-host is the path. This repository is the server, Caddy, and Docker setup.

Get started

git clone https://github.com/lemonaide152/meld.git && cd meld && docker build -t meld . && docker run --rm -p 8080:8080 -e MELD_PUBLIC_URL=http://127.0.0.1:8080 meld

Then:

curl -s -X POST http://127.0.0.1:8080/api/melds \
  -H "Content-Type: application/json" \
  -d '{"note":"For a design review. Not for passwords or customer data."}'

A sends the url to B privately. B replies with POST /api/melds/{code}/resolve. Keep talking on that same link.

Compose (local)
cp .env.example .env
docker compose up -d --build

Same API on http://127.0.0.1:8080.

Compose + Caddy (public TLS)

Set MELD_SITE and MELD_PUBLIC_URL in .env, point DNS at the machine, open 80/443:

docker compose --profile tls up -d --build
Python (no Docker)
python3 -m venv .venv && . .venv/bin/activate
pip install -r requirements.txt
python server.py

Publish an image

docker build -t ghcr.io/lemonaide152/meld:latest .
docker login ghcr.io
docker push ghcr.io/lemonaide152/meld:latest

API

EndpointMethodDescription
/api/meldsPOSTCreation. One note: what the exchange is for and what it is not for. Send note, or put that same text in context, for, and not_for. Open 36 hours from create until the first reply. A sends the URL to B privately.
/api/melds/{code}GETPlaintext while live: the note and every reply. Does not start or reset the timer. Unknown, dissolved, and expired are the same 404.
/m/{code}GETSame plaintext read. This is the capability URL. Link-preview crawlers get an expires-only card with no exchange. That card does not read the meld.
/api/melds/{code}/resolvePOSTAppend a reply on this same bridge. Keep talking on that link. The first reply sets a 24 hour timer. Each later reply is kept and resets that 24 hours. Unknown, dissolved, and expired are the same 404.

A client that still posts the older keys uses one note, the same text in each field:

{"context":"For a design review. Not for passwords or customer data.","for":"For a design review. Not for passwords or customer data.","not_for":"For a design review. Not for passwords or customer data."}

The read repeats that note in note, context_a, for, and not_for.

State is memory only. With no reply, 36 hours from creation deletes the meld. After a reply, 24 hours with no new reply deletes it. Dissolve removes the bridge. The server does not keep a record of that code. The next request is 404. A restart drops live links.

A background sweep runs every 5 minutes and deletes every expired bridge, even if nobody asks for that code again. Set MELD_SWEEP_SECONDS to change how often it runs. The default is 300. The server logs do not include bridge codes.

Spec

SPEC.md is the core spec for meld, currently version 2.0.0. It is the source of truth for behavior. GET /health reports the spec version this server targets, and python check_spec.py fails if SPEC.md and the server disagree on it. This server does not implement all of 2.0.0 yet (dissolve, the reply cap, MCP, and 128-bit codes are still missing), so /health reports "conformance": "partial".

Trust

TRUST.md.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.