General Tools
doorchannel.com
Provides general-purpose tools through the Model Context Protocol.
ENDPOINT 1
https://doorchannel.com/api/mcp
MCP server metadata
- Name
- doorchannel
- Version
- 1.0.0
DoorChannel is the buyer-side resting-order and escrow layer that agents route into. Money movement is TWO-PHASE: always call a prepare_* tool first, show the human the quote and the consent line verbatim, get their go-ahead, then call the matching send/place tool. Sending a knock COMMITS the saved card but charges nothing — the charge fires only when the recipient actually engages (a verified reply or a confirmed booking); silence is free. Credit orders charge when they clear the book. Credit orders bind to DoorChannel's versioned standard credit terms (no fees, 5+ years or no expiry, partial use). If a brand's final terms are worse, novel, or unreadable, the order enters pending_terms_confirm — NOTHING is charged; list_my_orders returns the plain-language diffs and an action_url. SURFACE that to your human and let them decide at the URL — never confirm or decline on their behalf, and never treat silence as consent (silence voids with zero charges). Privacy is absolute: you can knock on an identifier (email, LinkedIn, brand, person slug) but you can NEVER look one up. No response reveals who is behind a door, whether a door exists for an email, or links one person's doors. Order books are anonymized aggregates.
Known tools 11
get_brand_order_bookAnonymized order-book depth for a brand (by slug): the discount rungs and how many orders rest at each.
Inferred read-onlyget_multi_brand_clusterState of a multi-brand cluster (by key): the brand slate, member count, aggregate leading demand, and deadline.
Inferred read-onlyget_person_rate_cardA public person door by slug: the ways to reach them and their prices, plus public work history (company, title, start/end year or current) and the contact points the owner shows (email, X, ...).
Potential side effectsprepare_knockQuote a funded knock to a contact (email, LinkedIn URL, or person slug).
Potential side effectsget_knock_statusThe sender-visible status of one of your knocks: resting | delivered | accepted | released | refunded | expired, plus verdict_reason and standard_version once adjudicated — the published adjudication standard it was judged under, readable at github.com/DoorChannel/adjudication under the tag std-<standard_version>.
Inferred read-onlyCONNECT WITH APPROVAL
Client installation
Review this server and its permissions before adding it. Secret placeholders must be set locally.
Codex
~/.codex/config.toml
[mcp_servers.doorchannel]
url = "https://doorchannel.com/api/mcp"
enabled = true
Claude Code
.mcp.json
{
"mcpServers": {
"doorchannel": {
"type": "http",
"url": "https://doorchannel.com/api/mcp"
}
}
}
Claude Desktop
Settings → Connectors → Add custom connector
Name: doorchannel
Remote MCP URL: https://doorchannel.com/api/mcp
Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.
Cursor
.cursor/mcp.json
{
"mcpServers": {
"doorchannel": {
"url": "https://doorchannel.com/api/mcp"
}
}
}
Visual Studio Code
.vscode/mcp.json
Add to Visual Studio Code{
"servers": {
"doorchannel": {
"type": "http",
"url": "https://doorchannel.com/api/mcp"
}
}
}
Generic MCP
Client-specific MCP configuration
{
"name": "doorchannel",
"transport": "streamable-http",
"url": "https://doorchannel.com/api/mcp"
}
MCP Inspector
Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.
TRUST AND VERIFICATION EVIDENCE
Loading Trust v2 evidence…
Checking the associated registrable domain. The BuiltWith key remains server-side.
Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.