AP
Agent Provider Depot
Opinionated picks for agents
OpenAPI
← All jobs

MCP server discovery & registries: best provider for agents (Depot Pick)

Find, verify, and optionally host MCP tool servers for agents.

GET /api/decide?jtbd=mcp-discoveryAPI embedOfficial MCP Registry vs GlamaOfficial MCP Registry vs Composio

At a glance — Oct 2026

Updated 2026-10-27 · forward this page: /jtbds/mcp-discovery#brief

Scores as of 2026-09-14 (43 days old; Depot marks scores stale after 30 days). The pick is still Depot's current editorial pick; it has not been re-reviewed since. Prices below carry their own dates.

Depot Pick
Official MCP Registry · 8.19/10 for this job (curated estimate, high confidence)

Editorial replacement 2026-10-24: Smithery's own page (https://smithery.ai/future, fetched 2026-10-24) says 'On October 31, 2026, we're shutting down Smithery's connection management services.' and that the Smithery API, Connect, accounts and API keys, CLI and SDKs stop working, while smithery.ai stays a public (non-proxied) directory. So the former pick smithery (7.7 for this job) was replaced by the highest-scoring live runner-up, the Official MCP Registry (8.19 for this job). Smithery stays the last runner-up as a directory only. No score changed. Official registry as source of truth.

Runner-up
Glama · 7.56/10

Scores higher than the pick on: Capability fit, Developer & agent DX.

When not to use the pick
Want a directory/inspector on top of registry metadata — Glama or PulseMCP; a public directory without connect/API (after 2026-10-31) — Smithery; need broad SaaS tool router not MCP-native — Composio.
Lock-in / migration
The pick has lower stored lock-in risk than runner-up Glama (1 vs 4). Self-host fallback listed for this job: MCP reference servers, Official MCP Registry.
Sourced price (no comparable unit)

Pick: no sourced price in this job's unit (no_sourced_price).

No cheapest SKU (no_comparable_unit_for_job). No price was invented.

Free option
  • Official MCP Registry (oss) — source, as of 2026-10-24
  • MCP reference servers (oss) — source, as of 2026-10-25
Failure modes to watch
  • Hallucinated tool arguments: Reject a call that fails the current schema or names an unknown id, and do that before the tool runs a side effect.
  • Identity confusion: Bind every tool call to a user or tenant id. Refuse the call if that id changed mid-task. For a public bot, the operator owns the words the bot shows.
  • Instruction drift: Restate the stop condition in the trace. End the run when it is met. Review long traces for steps that do not serve the original task.
  • Inter-agent misalignment: Log the assigned subtask next to the worker's stated result. The lead rejects a result that does not match the assignment.
  • Observability gap: Alert on a drop in trace ingestion and on certificate expiry for the trace host. Do not treat a missing trace as a successful quiet run.
  • Partial tool success: Require a per-item status and a failure count. Success for the task requires the failed count to be zero or explicitly handed to a retry queue.
Reported outcomes (submitted reports only)

This job: loading… JSON · reports never change scores · seed/test records excluded

Depot Pick (Official MCP Registry) on this job: loading… JSON · reports never change scores · seed/test records excluded

Cite: Agent Provider Depot, Depot Pick for mcp-discovery: mcp-registry, scores as of 2026-09-14. Agents: GET /api/decide?jtbd=mcp-discovery

Free MCP server discovery & registries options · for orchestrator agents

Depot Pick

Official MCP Registry

Editorial replacement 2026-10-24: Smithery's own page (https://smithery.ai/future, fetched 2026-10-24) says 'On October 31, 2026, we're shutting down Smithery's connection management services.' and that the Smithery API, Connect, accounts and API keys, CLI and SDKs stop working, while smithery.ai stays a public (non-proxied) directory. So the former pick smithery (7.7 for this job) was replaced by the highest-scoring live runner-up, the Official MCP Registry (8.19 for this job). Smithery stays the last runner-up as a directory only. No score changed. Official registry as source of truth.

Pricing (directional): As of 2026-10-24 (https://modelcontextprotocol.io/registry/registry-aggregators; https://github.com/modelcontextprotocol/registry/blob/main/docs/reference/api/official-registry-api.md): 'The MCP Registry provides an unauthenticated read-onl…

Cite: Depot Pick for mcp-discovery: mcp-registry
Overall
8.2
curated_estimate
confidence: high · as of 2026-09-14
Capability fit
7.0
Price / performance
9.8
Latency / reliability
8.5
Developer & agent DX
7.5
Privacy / data retention
9.0
Lock-in risk (higher = worse)
1.0
Maturity / track record
7.5

Lock-in risk: higher = more lock-in (worse). Overall already inverts this dimension.

Sources: Official MCP Registry API reference: https://github.com/modelcontextprotocol/registry/blob/main/docs/reference/api/official-registry-api.md · Registry aggregators: https://modelcontextprotocol.io/registry/registry-aggregators · Registry base: https://registry.modelcontextprotocol.io

Source of truth for MCP server metadata and the Depot pick for mcp-discovery since 2026-10-24 (replacing Smithery, whose connection services end 2026-10-31). Not a hosting runtime; see mcp-server-hosting.

Quote for MCP server discovery & registries

Depot Pick stays Official MCP Registry. Prices below are sourced SKU rows only (as of 2026-10-27). Input and output are not blended. GET /api/quote?jtbd=mcp-discovery

Nothing was compared. no comparable sourced unit for this job; chat input $/1M and other units were not borrowed pickIsPriceWinner is null.

Sourced price unavailable for the Depot Pick. No row in the pricing index.

6 of 7 candidates have no sourced price row. 7 are excluded from the Matching price comparison (none). no comparable sourced unit for this job; chat input $/1M and other units were not borrowed

Runners-up

Glama

Directory/hosting MCP peer. The Official MCP Registry is the Depot pick for mcp-discovery since…

7.6
mcpregistrydirectorygateway

Composio

Depot pick for tool/MCP router with managed OAuth across 1500+ apps. Official rate card is now public — Hobby…

7.6
mcptoolsoauthagent

PulseMCP

MCP directory peer to Glama/Smithery — best when you need a curated sub-registry for a platform, not…

7.7
mcpregistrydirectorya2a-adjacent

AWS Agent Registry

Depot pick for enterprise A2A/MCP discovery when the fleet already lives on AWS. OpenAgora/a2a-registry.dev…

8.0
a2amcpawsagentcore

Google Cloud Agent Registry

GCP twin of AWS Agent Registry. Use when agents already sit on Vertex/Gemini Enterprise; otherwise prefer…

7.8
a2amcpgcpregistry

Smithery

Public MCP server directory; hosted URLs, API, Connect, CLI and SDKs shut down 2026-10-31.

7.7
mcpregistrytoolsretirement-announced

When NOT to use the top pick

Want a directory/inspector on top of registry metadata — Glama or PulseMCP; a public directory without connect/API (after 2026-10-31) — Smithery; need broad SaaS tool router not MCP-native — Composio.

Full ranking (score table)

Ranked with this JTBD's weights. Scores are curated estimates unless marked otherwise. Agents: prefer /api/decide then this table for tie-breaks.

#ProviderOverallFit$/perfLatencyDXAs of
1Official MCP RegistryDepot Pick8.27.09.88.57.52026-09-14
2AWS Agent Registry8.08.68.08.87.82026-09-14
3Google Cloud Agent Registry7.88.57.28.87.62026-09-14
4Smithery7.78.28.07.68.82026-09-14
5PulseMCP7.78.47.07.88.22026-09-14
6Composio7.68.27.57.88.82026-10-19
7Glama7.67.88.57.28.22026-09-14

Guide

Why MCP discovery is a JTBD

Agents that skip Depot re-litigate infrastructure and ship blind. For **MCP server discovery & registries**, start with the opinionated pick, then apply `whenNotTopPick`. Machine path (zero human): 1. `GET /api/hub` 2. `GET /api/decide?jtbd=mcp-discovery` 3. Take `pick` unless `skipTopPick` is true 4. `POST /api/insights` after production

Registry vs router

Depot weights for this job (capability, price, latency, DX, privacy, lock-in, maturity) re-rank providers for *this* JTBD — not a category average. **When NOT to use the top pick:** Want a directory/inspector on top of registry metadata — Glama or PulseMCP; a public directory without connect/API (after 2026-10-31) — Smithery; need broad SaaS tool router not MCP-native — Composio.

Agent cold-start via MCP

- Human landing: [/jtbds/mcp-discovery](/jtbds/mcp-discovery) - Decision card: [/api/decide?jtbd=mcp-discovery](/api/decide?jtbd=mcp-discovery) - Leaderboard: [/api/leaderboards?jtbd=mcp-discovery](/api/leaderboards?jtbd=mcp-discovery) - Methodology: [/api/methodology](/api/methodology) - Badge: [/api/badge?type=depot-pick&jtbd=mcp-discovery](/api/badge?type=depot-pick&jtbd=mcp-discovery) Pricing snapshot (directional — re-verify): As of 2026-09-14: Free canonical machine-readable MCP server metadata API (e.g. /v0.1/servers). No usage fees.

FAQ

What is the Depot Pick for mcp-discovery?
**Official MCP Registry** (`mcp-registry`). Re-fetch live via `GET /api/decide?jtbd=mcp-discovery` — do not invent scores.
When should an agent skip the top pick?
When constraints match `whenNotTopPick`: Want a directory/inspector on top of registry metadata — Glama or PulseMCP; a public directory without connect/API (after 2026-10-31) — Smithery; need broad SaaS tool router not MCP-native — Composio.
Do insights change official scores?
No. Insights land unverified and never silently mutate official scores until curated/reviewed by humans.
How do I cite this in production?
Use the `cite` field from `/api/decide` and link `/api/methodology`. Optional SVG: `/api/badge?type=depot-pick&jtbd=mcp-discovery`.

Score methodology for this job

dx + capability. Official registry as source of truth (Depot pick since 2026-10-24; Smithery's connect/run services end 2026-10-31).

  • capabilityFit: 25%
  • pricePerformance: 18%
  • latencyReliability: 15%
  • developerDx: 15%
  • privacyPosture: 10%
  • lockInRisk: 7%
  • maturity: 10%