canonical: https://jentic.com/apis/contrie.com/contrie # Contrie API The Contrie API fetches a web page and returns either accepted structured data or readable markdown, with source evidence and quality scoring attached. It can extract the same structured data or markdown from raw HTML you already hold, verify extracted values against the page they claim to come from, and search a curated reference index. Agents can also register a monitor that watches a single page for a single question and keep an eye on account credits. Each extraction carries signed evidence packets so a downstream reader can confirm where a value came from. ## For AI agents Fetch a URL or raw HTML and get back structured data or readable markdown with source evidence, verify extracted values against the page, and watch a page for a single question. ## Scope Does not handle whole-site crawling, headless browser automation, or proxy rotation. Use for single-page extraction, verification, and monitoring only. ## Capabilities - Fetch a URL and return structured data or readable markdown with quality scoring - Extract structured data or markdown from raw HTML you already hold - Verify extracted values against the page they claim to come from - Search a curated reference index for a question - Register a monitor that watches one page for one question - Check account credits and current restrictions before a run ## Use cases ### Agent-grade page reading An AI agent connected through Jentic calls the scrape operation to turn an arbitrary URL into structured data or markdown it can reason over, instead of loading a full HTML page into its context. Contrie attaches source evidence and a quality score so the agent knows how much to trust each field. Example prompt: Scrape a given product URL and return the title, price, and availability as structured fields with their evidence ### Extraction verification Teams that already scrape pages elsewhere use the verify operation to confirm that a value genuinely appears on the page it is attributed to, reducing silent drift. Contrie checks the extracted data against the live page and returns a verdict the pipeline can gate on. Example prompt: Verify that a previously extracted price value still appears on the source page and return the result ### Single-page change monitoring An agent registers a monitor that watches one page for one question, such as whether a job posting is still open, and reads back retained events over time. This keeps a lightweight watch without re-scraping the whole site on every check. Example prompt: Create a monitor for a careers page that watches whether a named role is still listed ## Key endpoints | Method | Path | Description | | --- | --- | --- | | POST | `/api/v1/scrape` | Fetch a URL and extract structured data or markdown | | POST | `/api/v1/extract` | Extract structured data or markdown from raw HTML | | POST | `/api/v1/verify` | Check extracted data against the source page | | POST | `/api/v1/search` | Search the curated reference index | | POST | `/api/v1/monitors` | Create a monitor that watches one page for one question | | GET | `/api/v1/account/status` | Check account credits and current restrictions | ## Key resources - **Scrape** — Fetch a URL and return structured data or markdown with evidence - **Extract** — Extract structured data or markdown from supplied raw HTML - **Verify** — Check extracted data against the page it claims to come from - **Monitors** — Create, list, read, and delete page monitors that watch one page for one question - **Account** — Read account credits and current restrictions ## AI readiness This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use. - **Score:** 54 / 100 - **Maturity:** Foundational - **Dimensions:** - Foundational Compliance: 48 / 100 - Developer Experience & Jentic Compatibility: 68 / 100 - AI-Readiness & Agent Experience: 39 / 100 - Agent Usability: 94 / 100 - Security: 60 / 100 - AI Discoverability: 41 / 100 - **View full report:** https://jentic.com/apis/contrie.com/contrie/scorecard - **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/ - **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview ### Score it yourself Every API in the directory is allowlisted, so you can re-score it with no key required. - **Score your own API:** https://jentic.com/scorecard.md - **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md ```sh npx @jentic/api-scorecard-cli score ``` ## Why Jentic - **Setup:** Wiring the Contrie API by hand means issuing a bearer token, learning which endpoints are public versus authenticated, and handling its credit and evidence model yourself. Through Jentic you install once, import Contrie from the API Directory, store the token once, and your agent calls it. - **Permission scoping:** Contrie's read operations carry their target in the request body, not the URL path, so rules bound which operations your agent may call, not which page. You choose the operations it may call, so it can scrape and extract without being able to create monitors or run the curated search unless you add those. - **Credential handling:** Your Contrie token is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context. - **Discovery method:** Agents search Jentic by intent such as 'extract data from a page' or 'verify a scraped value', and Jentic returns the matching Contrie operation with its input schema so the agent calls the right endpoint without browsing the reference docs. ## Related APIs - **Apify** — Apify runs full crawlers and browser actors, where Contrie focuses on single-page extraction with evidence. - **Diffbot** — Diffbot extracts structured article, product, and entity data from pages at scale. - **Exa** — Exa finds relevant URLs with neural search that you then read through Contrie. ## FAQ ### What authentication does the Contrie API use? The Contrie API authenticates with a bearer token sent in the Authorization header, per its OpenAPI spec, and a few endpoints such as evidence verification are public. Through Jentic the token is stored once by your own self-hosted instance and injected when a call runs, so it never appears in the agent's prompt or logs. ### Can I extract structured data from raw HTML with the Contrie API? Yes. Send the HTML to the extract operation and Contrie returns structured data or readable markdown without re-fetching the page. If you only have a URL, use the scrape operation instead, which fetches the page first and attaches source evidence. ### Can I limit what my agent is allowed to do with the Contrie API? Yes. Write a rule that allows only the scrape and extract operations, so the agent can read pages but cannot create monitors or spend credits on the curated search, and every call it makes is logged. Add more operations only when the agent needs them. ### What are the rate limits for the Contrie API? The spec documents per-plan limits of 10, 30, and 100 requests per minute per account across its keys on the Free, Builder, and Pro plans, along with monthly credit caps. Check the Contrie documentation for the current figures before running high-volume extraction. ### Is there a Contrie MCP server? You don't need an MCP server to give your agent the Contrie API. Jentic connects it directly from the API Directory: import it, store your token once, and your agent calls operations like scrape and verify on demand without loading another server's tool definitions into its context. ### How do I read a web page with the Contrie API through Jentic? Search Jentic by intent such as 'turn a web page into structured data', and Jentic returns the scrape operation with its input schema so your agent passes the URL and reads back structured fields and evidence. To run it on your own infrastructure, install Jentic One from its GitHub repo.