Skip to content

AgentWatchr MCP server

Which MCP servers are AgentWatchr's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.

read 2026-10-29

← Back to AgentWatchr

Whose server it is
  • Registry namecom.agentwatchr/evidence-indexverified in the MCP registry

The registry name sits under agentwatchr.com. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

When we asked itchecked 2026-10-29
  • Stateanswerschecked 2026-10-29
  • Tools4
  • Server calls itselfAgentWatchr
  • Server version0.1.0
  • Protocol version2025-03-26

On 2026-10-29 the server answered initialize and tools/list without a sign-in. First answer in 740 ms.

Should an agent connect itNot proven either way yet
  • Its tools do every use case we checked: look up an agent's recorded observations and their sources; find agents by name, identifier, or owner.
  • Checked on 5 of the 7 days it takes to judge reliability.

Once connected

  • For "find agents by name, identifier, or owner", the likely first call is search_agents.
  • Parameters are untyped or undescribed in 3 of 4 tools: check arguments before calling.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

Measured dailydaily checks: 5, the first on 2026-10-29
  • INITE MCP scoreafter a weeka score needs 7 daily checks: 5 so far
  • What an agent can do with it2 of 2 use cases
  • Answeredmeasuring5 of the 7 daily checks it takes to print a share
  • Median first answer1766 ms
  • Space in an agent's context≈ 204 tokensthe definitions of all its tools (4), which an agent carries on every turn
  • Tool definitionsC · 71 of 100
  • Typed results0 of 4 toolsdeclared outputSchema: results come typed, read once
  • Parameters without a description: in 3 of 4
  • Not saying whether it only reads or can destroy: 4 of 4

What an agent can do with it

  • can do · look up an agent's recorded observations and their sources
  • can do · find agents by name, identifier, or owner
  • not sure · pull agent evidence into other tools through the API

Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?

Worth a look

None of the five warning signs in these definitions.

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

Tools it offers4 tools · read 2026-10-29
  • search_agents

    Search observed source identities, not verified real-world agents.

  • get_agent

    Read profile, evidence and explained Evidence Score.

  • get_stats

    Read evidence and coverage distributions.

  • get_insights

    Read bounded FACT / INTERPRETATION / CONFIDENCE network analyses.

How to connect

Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:

{
  "mcpServers": {
    "evidence-index": {
      "type": "http",
      "url": "https://agentwatchr.com/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http agentwatchr https://agentwatchr.com/mcp

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://agentwatchr.com/mcp as a custom connector in the settings.

OpenAPI spec

12 paths, read 2026-10-29. Open the spec ↗

Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.