Skip to content
Vera logo

Vera MCP server

Which MCP servers are Vera'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 2026-10-26

← Back to Vera

Whose server it is
  • Registry nameink.vera/veraverified in the MCP registry

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

When we asked itchecked 2026-10-26
  • Stateanswerschecked 2026-10-26
  • Tools7
  • Server calls itselfvera-mcp
  • Server version1.2.1
  • Protocol version2025-06-18

On 2026-10-26 the server answered initialize and tools/list without a sign-in. First answer in 87 ms.

Should an agent connect itNot proven either way yet
  • Its tools do check supplier agreements against your procurement policy; flag clauses that break a cited statute or standard, but not draft corrected documents with tracked changes.
  • Checked on 1 of the 7 days it takes to judge reliability.

Once connected

  • For "check supplier agreements against your procurement policy", the likely first call is list_packs.
  • For "flag clauses that break a cited statute or standard", the likely first call is list_packs.

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: 1, the first on 2026-10-26
  • INITE MCP scoreafter a weeka score needs 7 daily checks: 1 so far
  • What an agent can do with it2 of 3 use cases
  • Answeredmeasuring1 of the 7 daily checks it takes to print a share
  • Median first answer75 ms
  • Space in an agent's context≈ 1,352 tokensthe definitions of all its tools (7), which an agent carries on every turn
  • Tool definitionsA · 100 of 100
  • Typed results7 of 7 toolsdeclared outputSchema: results come typed, read once
  • Every tool passes all six checks.

What an agent can do with it

  • can do · check supplier agreements against your procurement policy
  • can do · flag clauses that break a cited statute or standard
  • cannot · draft corrected documents with tracked changes

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 offers7 tools · read 2026-10-26
  • query_pack

    Retrieve the best matching locked rules and exact evidence from a query-ready Vera pack. This costs credits. Preserve each verified quote and citation, distinguish them from your inference, and treat coverage=none as no answer.

  • verify_document

    Start a paid, asynchronous check against a query-ready pack. Use one unique idempotency_key and reuse it only for an exact retry. Omit mode to use the pack's mode. Poll get_verification_status until terminal=true.

  • get_verification_status

    Poll a run using poll_after_ms until terminal=true. Only call it clean when verdict=clean, coverage_complete=true, and result_consistent=true; otherwise follow next_action.

  • list_packs

    List owned and public packs. Select from available_packs and require readiness.query_ready=true. If none fits, do not guess a pack_id: propose a finite pack plan, ask the user to approve a credit ceiling, then use start_pack_build if…

  • get_pack_coverage

    Inspect readiness, source coverage and freshness before high-stakes use. Disclose incomplete or stale coverage. query_ready means usable, minimum_coverage_ready means every measured section has a rule, and fully_strengthened means every…

  • start_pack_build

    Start a paid asynchronous pack build after the user explicitly approves max_credits. Supply exactly one of pack_id (an owned AI-buildable pack) or new_pack. The response returns immediately with job_id; poll get_pack_build_status only…

  • get_pack_build_status

    Check a durable pack-build job. Poll only while active=true. When stopped, use readiness_summary and present available_actions honestly; a usable budget pause offers use now or a separately approved finite continuation. Never restart…

How to connect

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

{
  "mcpServers": {
    "vera": {
      "type": "http",
      "url": "https://vera-mcp-production.up.railway.app/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http vera https://vera-mcp-production.up.railway.app/mcp

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://vera-mcp-production.up.railway.app/mcp as a custom connector in the settings.

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.