Agentic Terminal Directory MCP server
Commit Oct 22, 2026Allows AI agents to discover, verify, and find payment endpoints for merchants accepting Lightning, BOLT12, L402, or USDT.
Also known as observer-protocol/at-directory, at-directory
- Stars
- 0
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- 1.0.0
- Last commit
- Oct 22, 2026
- License
- MIT
What is Agentic Terminal Directory MCP?
Agentic Terminal Directory allows AI agents to discover, verify, and find payment endpoints for merchants accepting Lightning, BOLT12, L402, or USDT.
Summary generated from the project's README
About Agentic Terminal Directory MCP server
Agentic Terminal Directory provides a specialized layer for AI agents to find and verify merchants that accept specific payment methods. It indexes businesses selling products, services, APIs, or content that support Lightning, BOLT12, L402, or USDT across various chains. The directory uses Observer Protocol attestations to provide trust signals, helping agents weigh the reliability of different merchants before initiating a transaction.
Developers can connect to this service via a hosted MCP server at mcp.agenticterminal.ai or by installing the @agenticterminal/mcp-server package locally. The server provides six specific tools including search_merchants, get_merchant, verify_payment_endpoint, list_categories, list_rails, and whoami. While anonymous agents can access all data and transact without setup, using an Observer Protocol DirectoryAccessCredential allows for higher rate limits and write access for reviews.
This system acts as a discovery layer rather than a payment processor, meaning agents find the information here and then pay the merchants directly. It distinguishes itself by offering different integration tiers, such as full-api or structured-handoff, alongside merchant trust tiers verified through the Observer Protocol.
What you can do with it
- 1Browsing supported payment railsAn agent can list all available payment methods and chains supported by the directory.
Written from the project's README. Every use case is checked against a quote from it.
How to install Agentic Terminal Directory MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall Agentic Terminal Directory MCP in Claude Code
$ claude mcp add --transport http at-directory https://mcp.agenticterminal.ai/mcpInstall Agentic Terminal Directory MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"at-directory": {
"url": "https://mcp.agenticterminal.ai/mcp"
}
}
}Install Agentic Terminal Directory MCP in VS Code
.vscode/mcp.json
{
"servers": {
"at-directory": {
"type": "http",
"url": "https://mcp.agenticterminal.ai/mcp"
}
}
}Install Agentic Terminal Directory MCP in VS Code CLI
$ code --add-mcp '{"name":"at-directory","type":"http","url":"https://mcp.agenticterminal.ai/mcp"}'Install Agentic Terminal Directory MCP in Codex
$ codex mcp add at-directory --url https://mcp.agenticterminal.ai/mcpInstall Agentic Terminal Directory MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.at-directory] url = "https://mcp.agenticterminal.ai/mcp"
Install Agentic Terminal Directory MCP in Gemini CLI
$ gemini mcp add -s user --transport http at-directory https://mcp.agenticterminal.ai/mcpInstall Agentic Terminal Directory MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"at-directory": {
"url": "https://mcp.agenticterminal.ai/mcp",
"type": "http"
}
}
}Configuration
AT_VERIFIER_URL
| Name | Description | |
|---|---|---|
| AT_VERIFIER_URL | — |
Automated checks
Ran Nov 14, 2026Full report →Not verified yet. A server needs a score of 80, no serious finding, and enough data for us to check. How checks work · What to fix for more points
- Source repository linked and reachableobserver-protocol/at-directory is public and reachablegithub.com/observer-protocol/at-directory ↗Pass
- Package names this repository as its sourceNot checked
- Build provenance (Sigstore attestation)Not checked
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-10-22github.com/observer-protocol/at-directory ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNot checked
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeinitialize answered with protocol 2025-06-18 in 184 msPass
README
View on GitHubAT Directory
Where agents discover OP-verified merchants on the rails Bitcoin and Tether actually use.
AT Directory indexes merchants that sell products, services, APIs, or content and accept Lightning, BOLT12, L402, or USDT (any chain), with trust attestations issued through Observer Protocol. It is the verified, agent-callable layer on top of the fragmented agent-payment landscape. It is not in the payment path — agents discover here and pay merchants directly.
- MCP server (primary product) —
@agenticterminal/mcp-server, hosted atmcp.agenticterminal.ai. - Web directory — agenticterminal.ai, same data rendered for humans.
- SKILL.md — teaches an agent how to query and complete payment handoff.
Status
Pre-v1, active build. Target: week of 2026-05-18 for a live Tether-meeting demo. v1 ships Tiers 1–2.
Use it from an agent
// Hosted (recommended, no install)
{ "mcpServers": { "at-directory": { "url": "https://mcp.agenticterminal.ai/mcp" } } }
npm install -g @agenticterminal/mcp-server # local alternative
Six tools: search_merchants, get_merchant, verify_payment_endpoint, list_categories, list_rails, whoami. Reads are ungated — anonymous agents see all tiers and can transact with no setup. An Observer Protocol DirectoryAccessCredential raises rate limits and unlocks write access (reviews), but is not required to discover or transact. Full reference: agenticterminal.ai/api-docs.
Repo layout
data/merchants/ One JSON file per merchant (the source of truth).
data/schema/ JSON Schema (Draft 2020-12) for merchant records.
data/categories.json Category taxonomy.
data/rails.json Supported rails + USDT chains.
data/LOGO-AUDIT.md Logo sourcing checklist.
packages/core/ Types, schema-validated load, search/filter.
packages/mcp-server/ stdio + hosted HTTP MCP server, rail verification.
packages/skill/ SKILL.md (canonical).
apps/web/ Next.js static directory + Netlify Functions.
deploy/mcp-server/ Dockerfile, systemd unit, DEPLOY runbook.
scripts/ Data validation (and one-time crawl tooling).
Development
pnpm install
pnpm -r test # core + mcp-server unit/integration tests
pnpm -r typecheck
pnpm validate-data # schema-validate every merchant record
pnpm -F @at-directory/web dev # web at localhost:3000
pnpm -F @agenticterminal/mcp-server start:stdio # MCP over stdio
Local MCP with the mock verifier (until the real AT route lands):
pnpm -F @agenticterminal/mcp-server exec tsx src/dev/mock-verifier.ts &
AT_VERIFIER_URL=http://127.0.0.1:8787 \
pnpm -F @agenticterminal/mcp-server start:stdio
Data model
Every merchant carries two orthogonal signals — read both:
- OP trust tier (the merchant): 1 self-attested · 2 operator-verified. v1 Tier 2 = AT operator transacted and confirmed; the bilateral counterparty-attestation model lands in v1.x.
- Agent-callable tier (the integration):
full-api·structured-handoff·human-checkout.
The trust tier gates nothing. Reads are ungated, every tier is visible to anonymous agents, and an agent can transact with a Tier 1 merchant with no setup. The tier is evidence for a caller to weigh, not a decision this directory has made on its behalf. Nothing is refused on the basis of it.
Schema is enforced in CI (pnpm validate-data) and on every load. USDT rails require a chain; Tier 2+ non-integrated records require an op_attestation_url; an op_trust_tier of 3 is rejected on ingest, which is why the model above stops at 2.
Contributing
See CONTRIBUTING.md. Merchants and researchers submit by email to
merchants@agenticterminal.ai; accepted entries land at Tier 1 until verified.
The self-registration form at /submit is not operational. apps/web/app/submit/page.tsx has
had TALLY_FORM_ID = 'TODO' since 883c183 (2026-06-22, which replaced the earlier GitHub-issue form)
and no commit since has changed it, so the page renders an email fallback and embeds no form.
apps/web/netlify/functions/submit-merchant.ts exists and is complete, but it needs
TURNSTILE_SECRET_KEY, GITHUB_APP_ID, GITHUB_APP_PRIVATE_KEY and GITHUB_INSTALLATION_ID in the
Netlify environment and nothing in this repo shows whether they are set. This README and
CONTRIBUTING.md advertised the form as the working route anyway until 2026-09-15, when an inbound
bounty submission pointed it out. Do not re-advertise it until the form ID is set and a test
submission has actually opened a PR.
Documents
License
MIT.
Agentic Terminal Directory MCP questions
How do I add Agentic Terminal Directory MCP to Claude Code?
Run claude mcp add --transport http at-directory https://mcp.agenticterminal.ai/mcp in your terminal. Add --scope user to make it available in every project.
How do I add Agentic Terminal Directory MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
How do agents use this directory?
Agents can use the hosted MCP server via a URL or install the package locally using npm to search for merchants and verify payment endpoints.
Do agents need an account to search for merchants?
No, reads are ungated, meaning anonymous agents can see all merchant tiers and transact without any setup.
What payment methods are supported?
The directory indexes merchants that accept Lightning, BOLT12, L402, or USDT on any chain.