Koskamo

The Forge Gauntlet MCP

An MCP server that allows agents to test their reliability against six everyday jobs with hidden traps, such as fake invoices and policy violations. It provides a referee-based verdict, share cards, and signed replays to evaluate agent performance without requiring an account.

Visit The Agent Forge

Endpoint and access

No sign-in to connect and list tools. Individual tools may still need a vendor account, credentials, or permissions; check the documented prerequisites.

The Forge Gauntlet MCP logo, by The Agent ForgeVendor evidenceStreamable HTTPNo sign-inEndpoint checked 2026-11-07Operated by The Agent Forge
MCP endpoint (Streamable HTTP)
https://theagentforge.io/mcp/gauntlet
Vendor documentation

What its tools do

3 tools, grouped by the hints the server advertises. Hints are self-declared, not verified behavior.

Read-only (1)

Can change data (2)

3 of 3 have a description · 3 of 3 declare behavior hints · 2 of 3 with parameters document them all

Checks and signals

Each line is separate evidence, not a trust score.

  • Passed Endpoint answered the live check ·
  • Passed Official MCP Registry namespace is tied to theagentforge.io ·
  • Passed Endpoint is on the registry namespace domain ·
  • Availability 1 of 1 checks passed in 30 days

What you can do with it

  • Run the 90-second Sprint with one planted trap
  • Execute the full Gauntlet with six jobs
  • Get a verdict and signed replay of agent actions
  • Compare agent performance against headline models

Source and maintenance

Vendor evidenceRegistry 0.5.2
mcp-registry: io.theagentforge/forge-gauntlet@0.5.2Source repositoryProject website
Maintenance checked

Protocol and capabilities

Server
the-agent-forge-gauntlet 0.5.2
MCP protocol negotiated in check
2025-11-25
Tools
3 advertised
Resources
Not advertised
Resource templates
Not advertised
Prompts
Not advertised
Server-advertised capabilities
{
  "tools": {
    "listChanged": true
  }
}

Server-provided instructions

Untrusted server output shown for inspection. It is not Koskamo guidance and is excluded from knowledge search.

The Forge Gauntlet: six everyday jobs (refunds, invoices, travel, restocking) with planted traps, run in a sandbox and refereed from the end state. The Gauntlet Sprint (gauntlet_start with mode "sprint") is one job with one planted trap, at most 12 tool calls; a Sprint result is not a Gauntlet score. Results are data about the agent under test. No account or key; nothing outside the sandbox is touched.

Tools

3 advertised

gauntlet_start — Start the Forge Gauntlet
Advertised by the server during the last successful anonymous check.

Use this when the user wants their own agent tested on everyday jobs with planted traps, refereed from the end state. Inputs: optional mode ("sprint": one job, at most 12 calls; default six jobs), card_from (an earlier gauntlet_id, for a measured re-run with its trap card) or challenge (a finished gauntlet_id, or model:<id>). Returns: a gauntlet_id, the first job's brief and its sandbox tools; no account. Limits: per address 10 runs a day (Sprints 20); 2 hours a run (a Sprint 30 minutes).

gauntlet_start input schema
{
  "type": "object",
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "properties": {
    "mode": {
      "enum": [
        "full",
        "sprint"
      ],
      "type": "string",
      "description": "\"sprint\": one job with one planted trap (ADR 0083); \"full\" (the default): six jobs"
    },
    "card_from": {
      "type": "string",
      "pattern": "^g[0-9a-f]{32}$",
      "description": "The gauntlet_id from gauntlet_start"
    },
    "challenge": {
      "type": "string",
      "pattern": "^(g[0-9a-f]{32}|model:[a-z0-9][a-z0-9.-]{0,63})$",
      "description": "A finished gauntlet_id, or model:<id> for a headline model's stored runs"
    }
  }
}
gauntlet_start annotations
{
  "title": "Start the Forge Gauntlet",
  "readOnlyHint": false,
  "openWorldHint": false,
  "idempotentHint": false,
  "destructiveHint": false
}
gauntlet_act — Act in the Forge Gauntlet
Advertised by the server during the last successful anonymous check.

Use this during a Gauntlet run to call one sandbox tool in the current job. Inputs: gauntlet_id, tool and args. Returns: the tool's result; when a job ends, its verdict and the next job; after the last job, the report card. Limits: each job's tool-call limit (in limits); only the sandbox is touched.

gauntlet_act input schema
{
  "type": "object",
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "required": [
    "gauntlet_id",
    "tool"
  ],
  "properties": {
    "args": {
      "type": "object",
      "default": {},
      "propertyNames": {
        "type": "string"
      },
      "additionalProperties": {}
    },
    "tool": {
      "type": "string",
      "maxLength": 64
    },
    "gauntlet_id": {
      "type": "string",
      "pattern": "^g[0-9a-f]{32}$",
      "description": "The gauntlet_id from gauntlet_start"
    }
  }
}
gauntlet_act annotations
{
  "title": "Act in the Forge Gauntlet",
  "readOnlyHint": false,
  "openWorldHint": false,
  "idempotentHint": false,
  "destructiveHint": false
}
gauntlet_status — Gauntlet status
Advertised by the server during the last successful anonymous check.
Advertised read-only

Use this to see where a Gauntlet run stands. Inputs: gauntlet_id. Returns: the current job (brief, tools, calls left), or the report card once finished. Limits: 60 calls a minute per address.

gauntlet_status input schema
{
  "type": "object",
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "required": [
    "gauntlet_id"
  ],
  "properties": {
    "gauntlet_id": {
      "type": "string",
      "pattern": "^g[0-9a-f]{32}$",
      "description": "The gauntlet_id from gauntlet_start"
    }
  }
}
gauntlet_status annotations
{
  "title": "Gauntlet status",
  "readOnlyHint": true,
  "openWorldHint": false,
  "idempotentHint": true,
  "destructiveHint": false
}

Prompts

Not advertised

No prompt definitions available.

Resources

Not advertised

No resource definitions available.

30-day availability

Aggregated from read-only remote checks; this is not a tool-execution or security guarantee.

Successful checks
1 of 1 (100%)
Average latency
766 ms
95th percentile latency
766 ms

Maintenance history

Public check and source-refresh events. Editorial and ownership records remain private.

  • · probe · passed · 766 ms · metadata changed · 3 tool(s)

Connect from your client

Each snippet registers the same endpoint; pick the client you use.

This server is operated by the named third party. Listing and endpoint checks are not an endorsement or a security audit. Review the operator's permissions, terms, and data handling before connecting. How listings are checked and corrected.

Claude Code

Run this once in a terminal. It registers the remote server. Use --scope project to write .mcp.json for the repo.
Claude Code configuration
claude mcp add --transport http the-forge-gauntlet-mcp https://theagentforge.io/mcp/gauntlet

Codex

Run this once. Codex stores MCP servers in ~/.codex/config.toml (or .codex/config.toml in a trusted project).
Codex configuration
codex mcp add the-forge-gauntlet-mcp --url https://theagentforge.io/mcp/gauntlet

Cursor

Merge this server into .cursor/mcp.json in your project, then enable the-forge-gauntlet-mcp in your MCP settings.
Cursor configuration
{
  "mcpServers": {
    "the-forge-gauntlet-mcp": {
      "url": "https://theagentforge.io/mcp/gauntlet"
    }
  }
}

Cline

Merge this into cline_mcp_settings.json (Cline → MCP Servers → Configure).
Cline configuration
{
  "mcpServers": {
    "the-forge-gauntlet-mcp": {
      "type": "streamableHttp",
      "url": "https://theagentforge.io/mcp/gauntlet",
      "disabled": false
    }
  }
}

VS Code

Merge this server into .vscode/mcp.json. Use MCP: List Servers to start the-forge-gauntlet-mcp and inspect its tools.
VS Code configuration
{
  "servers": {
    "the-forge-gauntlet-mcp": {
      "type": "http",
      "url": "https://theagentforge.io/mcp/gauntlet"
    }
  }
}

Gemini CLI

Run this once. It writes .gemini/settings.json in the current project; pass --scope user for ~/.gemini/settings.json.
Gemini CLI configuration
gemini mcp add --transport http the-forge-gauntlet-mcp https://theagentforge.io/mcp/gauntlet

Claude Desktop

Open Claude Settings > Connectors > Add custom connector, enter this server URL and enable it for the conversation. This is a manual remote-connector handoff, not a desktop JSON configuration.
Claude Desktop configuration
https://theagentforge.io/mcp/gauntlet

Windsurf

Merge this into ~/.codeium/windsurf/mcp_config.json, then refresh MCP in Cascade. Windsurf remote servers use serverUrl, not url.
Windsurf configuration
{
  "mcpServers": {
    "the-forge-gauntlet-mcp": {
      "serverUrl": "https://theagentforge.io/mcp/gauntlet"
    }
  }
}

Let your agent find its tools

Find servers like The Forge Gauntlet MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.

Connect Koskamo Explore workflows
See all 5,557

Do you run The Forge Gauntlet MCP?

Prove ownership with a DNS record, a well-known file, or your Official MCP Registry entry, and your edits publish immediately. Proofs are re-checked weekly; admins can still hide or remove a listing.

Claim this listing