22 Services from shelf.thirdmade.net

Analyzes a token contract address for honeypot signals, proxy detection, and liquidity pool data to help trading agents avoid rug pulls and scams.

Scores a given text input for prompt-injection and jailbreak risk so agent builders can sanitize user inputs before forwarding them to an LLM.

Scrapes any public web page and returns clean structured JSON containing the page title, extracted text, and content length.

Generates a 3-sentence extractive summary and top-5 keywords from any input text.

Extracts and classifies all outbound links from a webpage, grouped by domain, showing where a page links and which domains dominate.

Checks whether a given URL supports or substantiates a specific claim, returning a confidence score for source verification.

Generates a single 512px PNG image from a text prompt using FLUX.1-dev on GPU-backed RunPod ComfyUI, returning a job ID and poll URL for async retrieval.

Detects personally identifiable information (emails, phone numbers, SSNs, credit card numbers) in text and returns character-level span positions for each match.

Validates a JSON string for correctness, detects PII within its values, and reports schema statistics for use as a data-quality gate.

Returns the current transaction nonce for any EVM address on Base or Ethereum, enabling agents to construct valid transactions.

Resolves an ENS name to an Ethereum address (or vice versa) using the public Ethereum RPC.

Live-checks a domain for reachability, TLS/HTTPS status, and page title in a single GET request.

Returns geocode coordinates, timezone, and currency context for any free-text location string in a single GET request.

Converts messy, unstructured text (key=value pairs, CSV-like lines, free sentences) into clean, structured JSON

Fetches and extracts OG tags, canonical URL, title, and meta description from any public web page URL.

Scores whether a provided context passage entails or supports a given claim, helping research agents self-verify their outputs.

Extracts named entities (people, companies, and places) from arbitrary input text via a lightweight NLP probe.

Classifies input text as positive, negative, or neutral sentiment and returns matching emotion keyword hits.

Returns current block height and chain statistics for Base, Ethereum, or Arbitrum.

Finds similar and competitor domains by name-similarity for a given domain input.

Identifies the type and difficulty of a CAPTCHA on a given page (reCAPTCHA, hCaptcha, Turnstile, image, slider, puzzle) — evaluation only, no solving.

Runs a single verify-loop research sweep on any topic, returning confidence-labeled findings