Skip to main content
Glama
intesta-mcp

intesta-mcp

intesta-mcp

An MCP server that lets an AI agent check who it is dealing with before it acts.

Intesta is a trust registry. Every entity has a trust level on the A0→A4 ladder (A0 claimed · A1 domain-control · A2 payment-verified · A3/A4 higher) and a set of facts, each marked confirmed (independently verified) or auto_collected (scraped from public sources, not verified). This package is a thin proxy over the public HTTP API and exposes the same tool names and shapes as the remote server https://intesta.io/mcp — one contract on every surface (since 0.2.0).

Its ethos is Intesta's own: refuse, don't guess. A level is only ever what was proven; an auto_collected fact is not a verified one. The server reports exactly what the registry says and never upgrades either.

Tools

Tool

What it answers

check_trust(domain, purpose="payment")

One call, one decision, any domain. known, level A0..A4, attested (true only at A1+), verdict (sufficient / insufficient / unknown) under a named policy profile (payment/v1 needs A2, data_share/v1 and follow_instructions/v1 need A1), what_is_proven, proof, expires_at, next_step, signature. Unlisted domain → unknown, never "failed".

get_passport(entity_domain)

The full fact passport: facts with attestation_status / evidence_class, as_of, effective_expiry, stale, plus the attestation block, passport_hash and signature.

ask(entity_domain, question)

A question answered only from that entity's facts, signed (passport_hash + registry signature). A refusal means the passport has no such fact.

search_entities(query, limit=20)

Find entities by name / domain / vertical (literal substring, ≥3 chars); each result carries its level.

list_entities()

The registry (first 500 + total).

get_trust_ladder()

What each of A0..A4 means and how to use the registry as an agent.

Deprecated 0.1.x aliases (search_registry, get_facts, ask_entity, trust_ladder) still work for one release and are removed in 0.3.0.

All tools are read-only and need no API key — public registry reads work out of the box. An optional INTESTA_API_KEY (ik_…) is only needed for metered or private use.

Errors (unknown entity, invalid question or query, API failure) are returned as MCP error results (isError: true), never as data-shaped successes.

Related MCP server: a2a

Install

// Claude Desktop / any MCP client — mcp config
{
  "mcpServers": {
    "intesta": { "command": "uvx", "args": ["intesta-mcp"] }
  }
}

Or connect to the remote server directly (same tools, no install): https://intesta.io/mcp (streamable-http).

Example

check_trust("acme-invoicing.com")
→ {"known": false, "level": null, "attested": false, "status": "not_checked",
   "verdict": "unknown", "policy_profile": "payment/v1",
   "next_step": "Not in the Intesta registry: nothing is proven either way. Treat as unverified. ...",
   "signature": {...}}

check_trust("darkpanther.eu", purpose="data_share")
→ {"known": true, "level": "A1", "attested": true, "verdict": "sufficient",
   "what_is_proven": "...", "proof": {"method": "well_known", "verified_at": "..."},
   "expires_at": "2026-10-03T...", "signature": {...}}

A verdict is the registry's proof measured against a published profile — never a guarantee, never financial advice. Verify the signature against https://intesta.io/.well-known/intesta-key (see https://intesta.io/llms.txt).

Development

pip install -e ".[test]"
pytest

License: MIT. Publisher: DC ESCRYPT SL — https://intesta.io

Available Tools

5 tools
ask_entityA

Ask a question answered ONLY from a specific entity's registry facts. The answer is grounded in that entity's passport (no guessing) and carries a passport_hash and, when configured, a registry signature you can verify offline. Use for 'what does X do / what is X's refund policy' about a known domain.

Inputs: domain (required), question (required). Returns: result

check_trustA

Check what is actually proven about a domain before you trust it. Returns its trust level (A0..A4), whether it is attested, the human-readable 'what_is_proven', the proof method, and how many public facts it has. A low level or attested=false means unproven — not necessarily false, but not verified.

Inputs: domain (required). Returns: result

get_factsA

Read a domain's public facts from the registry. Each fact carries an attestation_status: 'confirmed' (independently verified) or 'auto_collected' (scraped from public sources, NOT verified). Do not treat auto_collected as proven. Also returns each fact's source so you can check it yourself.

Inputs: domain (required). Returns: result

search_registryA

Search the Intesta trust registry for entities by name or domain. Returns each match with its domain, trust level (A0..A4) and public fact count. Use this to find the domain to pass to check_trust / get_facts.

Inputs: query (required), limit. Returns: result

trust_ladderA

Explain the A0..A4 trust ladder: what each level means and what it proves. Call this to interpret the 'level' returned by check_trust and search_registry.

Inputs: no named parameters. Returns: result

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.9/5.0

    Scored across 5 tools

    Disambiguation5/5

    Each tool targets a distinct step in the trust workflow: discovery via search_registry, trust assessment via check_trust, fact retrieval via get_facts, grounded Q&A via ask_entity, and interpretation via trust_ladder. Boundaries are clear, and descriptions explicitly guide when to use each tool.

    Naming Consistency4/5

    Most names use a consistent verb_noun snake_case pattern (search_registry, check_trust, get_facts, ask_entity). The one outlier is trust_ladder, which is noun-based rather than verb-based, but the overall convention remains readable and predictable.

    Tool Count5/5

    Five tools is well-scoped for a trust registry lookup server. Each tool has a clear purpose and no unnecessary redundancy, supporting a focused read-oriented workflow.

    Completeness4/5

    The toolset covers the core lifecycle: find entities, check trust, read facts, ask grounded questions, and explain the trust ladder. A minor gap is the lack of an explicit tool to verify returned passport signatures or hashes offline, though the descriptions suggest this can be done externally.

    Maintenance

    ActivityStale
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI agents to get read-only trust verdicts on knowledge records by evaluating evidence maturity, provenance, intended use, risk level, and scope before deciding whether knowledge may be used.
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables AI agents to check trustworthiness before recommending URLs, products, or organizations, with fail-closed pass/fail verdicts and attested-only recommendations.
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI agents to safely inspect a local repository's code and metadata while blocking private data from leaving the machine, providing read-only tools for search, change tracking, and integrity verification.
      MIT
    • A
      license
      A
      quality
      B
      maintenance
      Enables AI assistants and agents to perform read-only lookups of dated on-record facts about agents, wallets, and endpoints, including evidence levels and history, so they can assess what is known before paying or relying on them.
      4
      121 npm
      MIT