---
name: netintel-production-440c.up.railway.app
description: This host exposes a single skill that parses unstructured or semi-structured text into a typed JSON object using a caller-supplied JSON Schema. It uses an LLM-backed extraction pipeline and reports token consumption alongside the extracted result. No schema design, generation, or binary input processing is supported.
host: netintel-production-440c.up.railway.app
---

# netintel-production-440c.up.railway.app

netintel-production-440c.up.railway.app is a narrow-purpose text extraction service. It accepts raw text and a JSON Schema, then returns a structured object whose fields are defined entirely by the caller's schema. It is aimed at agents that already know the shape of the data they want and need to pull it out of free-form text such as scraped web content, API prose responses, or document snippets.

## When to use this host

Use this host when an agent has raw text and a known target schema and needs to extract typed, structured data from that text in a single call. It is well-suited for parsing contact records, contract fields, or API response prose into a predictable JSON shape. Do not use it for designing or generating schemas — the caller must supply the schema. Do not use it for binary inputs (PDFs, images, audio) or for tasks that require multi-step reasoning, summarization, or transformation beyond field extraction. For schema generation or document understanding pipelines that include OCR or binary parsing, a different host with those capabilities is required.

## Capabilities

### Structured Text Extraction

Maps free-form text onto a caller-defined JSON Schema, returning only the fields specified in that schema along with token usage metadata.

- **`extract-structured-schema-from-text`** — Extracts structured JSON data from raw text by mapping it to a caller-supplied JSON Schema, returning only the fields defined in the schema.

## Skill reference

### `extract-structured-schema-from-text`

**Schema Extractor** — Extracts structured JSON data from raw text by mapping it to a caller-supplied JSON Schema, returning only the fields defined in the schema.

*Use when:* Use when an agent needs to parse unstructured or semi-structured text (e.g. contact info, API responses, contract artifacts) into a typed JSON object conforming to a known schema.

*Not for:* Do not use for generating or designing new schemas; this endpoint extracts data into an existing schema, not creates one. Not suitable for binary or non-text inputs.

**Inputs:**

- `raw_text` (string, required) — The unstructured or semi-structured text from which data should be extracted.
- `target_schema` (object, required) — A JSON Schema object defining the structure and types of fields to extract from raw_text.

**Returns:** Returns an 'extracted' object with schema-matched fields populated from the raw text, plus a 'tokens_used' object with input and output token counts.

**Example:** `{"raw_text":"Hi I'm Sarah Chen, VP of Engineering at Acme Corp. Reach me at sarah.chen@acme.com or 555-867-5309.","target_schema":{"type":"object","required":["name","email"],"properties":{"name":{"type":"string"},"role":{"type":"string"},"email":{"type":"string"},"phone":{"type":"string"},"company":{"type":"string"}}}}`

---