Skip to main content
Glama
mynameistito Agent Directory

mynameistito Agent Directory

by mynameistito

mynameistito Agent Directory

CI License: MIT

A small, publisher-maintained catalog of mynameistito products and their documented interfaces. The directory is available as JSON and through a read-only Model Context Protocol (MCP) server using Streamable HTTP.

The catalog describes products; it does not execute or proxy them. Entries are not independent security, identity, or availability attestations.

Endpoints

Related MCP server: mcp-server-reference

MCP capabilities

Connect an MCP client to https://agents.mynameistito.com/mcp using the Streamable HTTP transport. The server is stateless and read-only.

Tools

Tool

Input

Behavior

list_products

Optional protocol: mcp, rest, or cli

List catalog entries, optionally filtered by one of their actual interfaces.

get_product

Required stable product id

Return one product with its published interfaces and limitations; returns null if no ID matches.

search_products

Required query; optional protocol

Search IDs, names, descriptions, and limitations, optionally filtering by interface.

Tool results include both readable JSON text and structured content. They describe entries only and never make calls to a listed product.

Resources

  • directory://manifest — complete versioned catalog JSON.

  • directory://products/{id} — a single product entry, or a not-found result.

Prompt

  • choose_product_for_task(task) — asks a client to compare a task with current catalog entries, state interface fit and limitations, and avoid inventing unsupported capabilities.

Current entries

  • mynameistito Agent Directory — this MCP catalog and its JSON manifest.

  • Hamilton Bin Day — REST lookup API, @mynameistito/hcc-bin-day CLI, and a read-only MCP lookup endpoint at https://bin-day.mynameistito.com/api/mcp. It is not described as an A2A service.

Update src/domain/catalog.ts when adding products. Add only interfaces that exist and are publicly documented. Keep descriptions and limitations factual; a DNSSEC-signed DNS record or a publisher-controlled catalog is not a security audit or guarantee of an endpoint's behavior.

DNS-AID

DNS-AID is currently an Internet-Draft and its index document format is not standardized. The public DNS zone mynameistito.com is DNSSEC-signed. This organization-index entry points to the production directory:

_index._agents.mynameistito.com. 3600 IN SVCB 1 agents.mynameistito.com. alpn="h2,http/1.1" port=443

This points discovery to the HTTPS organization directory; the MCP endpoint itself is https://agents.mynameistito.com/mcp. It advertises only HTTP transports (h2 and http/1.1), not a draft-only mcp ALPN identifier or an invented experimental SvcParamKey. Check the current DNS-AID draft and RFC 9460 before changing experimental SVCB parameters.

The directory is also published at the service hostname agents.mynameistito.com. DNS-AID scanners that treat that hostname as the organization domain query its own well-known owner name, so this second pointer is published as well:

_index._agents.agents.mynameistito.com. 3600 IN SVCB 1 agents.mynameistito.com. alpn="h2,http/1.1" port=443

The Hamilton Bin Day service is published at bin-day.mynameistito.com. Since discovery checks may use that service hostname as the organization domain, it also points to the central directory:

_index._agents.bin-day.mynameistito.com. 3600 IN SVCB 1 agents.mynameistito.com. alpn="h2,http/1.1" port=443

All three records are ServiceMode SVCB pointers to the central directory and are covered by the DNSSEC-signed mynameistito.com zone. alchemy.run.ts declares the pointers and DNSSEC for the production stage. Existing records and the zone are explicitly adopted so Alchemy can manage them without replacing them.

Development

Requires Bun 1.4.2. GitHub CI runs on ubuntu-24.04. The project has no Node.js runtime dependency: Bun is used for development, tests, builds, and deployment. GitHub JavaScript Actions run on Node.js 24; Cloudflare's nodejs_compat is a Worker API compatibility flag, not a Node version pin.

bun install
bun run dev
bun run check
bun run typecheck
bun run test
bun run build

Alchemy owns Cloudflare Worker deployment and state:

$env:STAGE = "prod"
bun run deploy

Preview stages omit the production custom domain. The production stage binds agents.mynameistito.com.

GitHub deployment

.github/workflows/deploy.yml uses mynameistito/alchemy-deploy, pinned to the immutable v4.0.2 commit. The CI workflow builds and uploads a credential-free Worker artifact; production deploy and PR previews are performed by the trusted reusable action.

Configure this repository variable and secret before deployment:

  • CLOUDFLARE_ACCOUNT_ID (Actions variable)

  • CLOUDFLARE_API_TOKEN — narrowly scoped to the Cloudflare account and required Worker deployment operations.

For the API token, grant Account → Workers Scripts → Edit on the target account, plus Zone → Workers Routes → Edit, Zone → DNS → Edit, and Zone → Zone → Read for mynameistito.com. DNS Edit is needed for Alchemy to adopt and reconcile the SVCB records and DNSSEC state. Cloudflare provisions the Custom Domain's DNS and certificate. Use a Cloudflare API token, not the global API key.

PR builds receive no Cloudflare credentials. Fork PRs do not receive privileged previews. Scheduled reconciliation cleans up previews whose PR-close event was missed.

License

MIT. See LICENSE.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers