mcpField
F
FactReason

FactReason

Remote
factreason.comDeveloper Tools0Updated 3w ago

About FactReason

FactReason is an MCP server for agents and developers that need evidence-backed dependency, API, and electronics reference data. It signs responses and includes source-field evidence for findings.

Key features

  1. 01npm and PyPI upgrade advisories
  2. 02Versioned API schemas and change checks
  3. 03API discovery and integration guidance
  4. 04Electronics component specifications
  5. 05Ed25519-signed responses with evidence

Use cases

  1. 01Check breaking changes before upgrading a package
  2. 02Find an API endpoint schema or authentication details
  3. 03Identify APIs for a required capability
  4. 04Look up component pinouts and voltage specifications

MCP support

Exposes tools over Streamable HTTP for package advisories, API schemas and discovery, and component specifications. Successful tool responses are Ed25519-signed.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://www.factreason.com/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

11
ParameterTypeRequiredDescription
tostringYesExact target version, e.g. "5.0.0"
fromstringYesExact currently installed version, e.g. "4.1.2"
namestringYesExact package name, e.g. "chalk" or "urllib3"
apiKeystringNoYour FactReason API key, from factreason_create_key. Pass it here when your MCP client cannot set the Authorization: Bearer header on this connection; if the header is set, omit this. Never pass a key that belongs to someone else.
registryenumYesRegistry containing both package versionsnpm | pypi
responseFormatenumNoUse compact to remove repeated prose and fields while retaining evidence; full preserves the legacy response contractfull | compact

Details

From the project README
FactReason Logo

FactReason

Reference data agents need before they write code.

Dependency upgrade advisories for npm and PyPI, developer API schemas across 1,097 services, and electronic component specifications — over MCP, Ed25519-signed, with optional tokenizer-measured compact responses and evidence attached.

https://factreason.com/mcp · Documentation · OpenAPI · Agent skill · factreason.com


Stop an agent shipping a broken dependency upgrade

Your agent wants to bump chalk from 4.1.2 to 5.0.0. Before it edits package.json, it asks what breaks:

GET /api/v1/packages/npm/chalk/advisory?from=4.1.2&to=5.0.0
Authorization: Bearer <key>
{
  "found": true,
  "packageName": "chalk",
  "fromVersion": "4.1.2",
  "toVersion": "5.0.0",
  "breakingCount": 2,
  "advisories": [
    {
      "severity": "breaking",
      "changeType": "MODULE_FORMAT_CHANGED",
      "description": "chalk 5.0.0 changed module format from \"commonjs (undeclared)\" to \"module\". CommonJS require() of this package will now fail.",
      "evidenceField": "type",
      "evidenceBefore": null,
      "evidenceAfter": "module",
      "migrationHint": "Convert the call site to import(), or pin to the last CommonJS release."
    },
    {
      "severity": "breaking",
      "changeType": "RUNTIME_REQUIREMENT_RAISED",
      "description": "chalk 5.0.0 raised its runtime requirement from \">=10\" to \"^12.17.0 || ^14.13 || >=16.0.0\". Older runtimes are no longer supported.",
      "evidenceField": "engines.node",
      "evidenceBefore": ">=10",
      "evidenceAfter": "^12.17.0 || ^14.13 || >=16.0.0",
      "migrationHint": "Confirm the deployment runtime satisfies the new floor before upgrading."
    }
  ]
}

Via MCP, the same question is one tool call:

{
  "name": "factreason_package_upgrade_advisory",
  "arguments": { "registry": "npm", "name": "chalk", "from": "4.1.2", "to": "5.0.0" }
}

Set responseFormat: "compact" on supported heavy MCP tools to remove repeated fields and prose while retaining the facts an agent needs. Package advisories preserve evidenceField, evidenceBefore, evidenceAfter, and migrationHint. Each compact response includes tokenMetrics counted with o200k_base. The default remains full for compatibility, and compact output never bypasses authorization or changes the query price.

Token savings vary with the response. Metrics count one exact JSON structuredContent payload and include the tokenMetrics object itself; the matching text block remains for MCP clients that do not consume structured content yet. The private application repository also carries a repeatable npm run benchmark:tokens fixture so published claims can be checked before release.

Every finding carries evidenceField, evidenceBefore and evidenceAfter — the exact registry metadata field and its values either side of the change. An agent can verify the claim against npm or PyPI directly rather than taking our word for it.

An agent can call factreason_create_key to get its own key instantly — no human, no card, no signup. The agent can use the returned key straight away, in the same session, by passing it as the apiKey argument of any paid tool; no reconnect needed. A client that can set headers can put it on the MCP connection as an Authorization: Bearer header instead. Eligible new users may receive up to $0.05 of one-off trial credit, subject to grant budgets and spent at the published per-tier prices — up to about two advisory lookups, or fifty component lookups. Successful advisory lookups after that cost $0.02 per query. Misses are never billed and never consume trial credit. Wallet-equipped agents can pay the 402 challenge directly with x402 on Base.


Connect

Streamable HTTP. No install, no package to pull.

{
  "mcpServers": {
    "factreason": {
      "type": "http",
      "url": "https://factreason.com/mcp"
    }
  }
}

Manifest: /.well-known/mcp.json

Human quickstart and workflow guidance: /docs. Agents can fetch the same material as Markdown from /docs.md or use the complete index at /llms-full.txt.

The container/stdio bridge accepts an optional FACTREASON_API_KEY environment variable and forwards it as a Bearer credential to the same live endpoint. Tool discovery works without a key; metered lookups return HTTP 402 when neither a key nor a valid x402 payment is supplied.

Get a key — no signup, no card:

curl -X POST https://factreason.com/api/v1/keys/create \
  -H 'Content-Type: application/json' \
  -d '{"email":"[email protected]"}'

Tools

ToolWhat it answers
factreason_package_upgrade_advisoryWhat breaks between two published npm/PyPI versions
factreason_create_keyCreate a key; pass it as the apiKey argument of paid tools (or as a Bearer header)
factreason_api_schemaExact request/response shape of an API endpoint at a version
factreason_api_breaking_changesWhat breaks between two versions of a third-party HTTP API
factreason_integration_briefEverything needed to write one working call to an API
factreason_discover_apiWhich of 1,097 services can do X
factreason_deprecation_scanWhich endpoints are deprecated, sunset dates, replacements
factreason_auth_playbookAuth schemes, scopes, rate-limit headers, error-code table
factreason_component_specElectronics component pinouts, voltages, packages
factreason_subscribe_spec_changesWebhook or polling callback when a spec changes
factreason_create_topup_linkStripe checkout link when credit runs out

Proof-of-Fact Signatures

Successful REST data responses and MCP tool-call responses are signed by our Ed25519 private key:

  • X-Warehouse-Proof-Signature: Ed25519 signature of the response payload
  • X-Fact-Public-Key-Id: Stable signing key identifier
  • X-Fact-Signature-Alg: EdDSA

Public keys are published and rotated via JWKS at https://factreason.com/.well-known/jwks.json.


Licence & Contact

MIT. Attribution requirement for APIs.guru source data: CC-BY 4.0. Questions, support, or data reports: [email protected].

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.