Claix

Extract PDF to JSON

claix.extract.pdf

Read-onlyIdempotent

Extract structured JSON from a PDF (text or scanned, max 15 MB) using a Claix pdf-json schema. Returns a single JSON object in data[]. Use for invoices, contracts, forms, or reports.

Input Schema

TableJSON Schema

Name	Required	Description
`api_key`	No	API key secreta de Claix. Ejemplo: claix_sk_abc123.... Opcional si la conexión MCP envía x-api-key en cabecera HTTP (recomendado en Smithery/Cursor).
`filename`	No	Nombre original del archivo con extensión. Ejemplo: factura-2026-03.pdf. Ayuda a inferir el MIME cuando envías file_base64.
`file_path`	No	URL HTTPS pública del archivo a procesar. Ejemplo: https://cdn.example.com/factura.pdf. No uses rutas locales del PC del usuario.
`schema_id`	Yes	UUID del schema creado en el dashboard de Claix. Ejemplo: 550e8400-e29b-41d4-a716-446655440000. Llama a claix.schemas.list primero si no lo conoces.
`file_base64`	No	Archivo codificado en Base64. Acepta data URLs (data:application/pdf;base64,...) o Base64 puro. Ejemplo de uso: adjunta el PDF/imagen del chat como Base64 antes de llamar a extract_*.

Output Schema

TableJSON Schema

Name	Required	Description
`data`	No	JSON payload from the Claix API (extracted records, schema list, or Excel export metadata).
`error`	No	Human-readable error message when success is false.
`success`	Yes	True when Claix returned a successful response. False when isError is set on the tool result.

TDQS

A4.2/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint, idempotentHint, and non-destructive behavior. The description adds valuable context beyond this: file size limit (15 MB), support for scanned PDFs (OCR), and the exact output location (data[]). No contradictions. This is a good but not exhaustive disclosure (e.g., no rate limits or output schema explanation).

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two tight sentences with no filler. The first sentence front-loads the verb, resource, and key constraints; the second sentence provides use cases. Every word earns its place.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a medium-complexity tool, the description covers the essential purpose, output format, size limits, and input type. An output schema exists and annotations are present, so the description need not explain return values or safety. A minor gap: the need to create/clarify a schema beforehand is only in the parameter description, not the main description, but it is still discoverable.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has 100% description coverage, so the baseline is 3. The main description does not add much parameter-level detail; it mentions 'using a Claix pdf-json schema' which relates to schema_id, but the schema descriptions already explain all five parameters clearly. No need for extra compensation.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description uses a specific verb ('Extract') and identifies the resource ('PDF') and output ('structured JSON'), with a clear scope ('text or scanned, max 15 MB'). It also names the schema type ('Claix pdf-json') and output shape ('single JSON object in data[]'), which differentiates it from sibling extract tools for doc/excel/image.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides clear use cases ('invoices, contracts, forms, or reports') that indicate when to apply this tool. It does not explicitly mention alternatives or exclusions, but the sibling tool names (e.g., claix.extract.doc) make the differentiation implicit. This meets 'clear context, no exclusions'.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All Claix tools Inspect

Try in Browser

RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

TDQS

A3.9/5.0

Disambiguation5/5

Every tool has a clearly distinct role: extract.* produces raw structured JSON, agent.* adds agent-mode reasoning, schemas.* manages schemas, and window_context.* handles persisted documents. Even the parallel extract/agent pairs for each format are disambiguated by the agent/group prefix and explicit descriptions about is_agent_mode.

Naming Consistency4/5

All tools follow a claix.<group>.<target> convention with lowercase snake_case, which is predictable and readable. The main inconsistency is action placement: extract.doc is verb-first while schemas.create is object-first, and convert.json_to_excel uses a noun phrase instead of a verb.

Tool Count4/5

At 17 tools, the set is slightly above the ideal 3-15 range, but each tool maps to a distinct endpoint or format variant. The parallel extract and agent families are justified by different processing modes, though they do make the surface feel a bit heavier than necessary.

Completeness4/5

The set covers schema lifecycle (create, list, delete), extraction for five major formats, agent-mode variants, document deletion, and window-context query/retrieval. Minor gaps include no schema update endpoint and no generic document listing/retrieval outside window_context.

Resources

Need Help?