All projects matching “agentic”

Deterministic prompt-injection firewall and honeypot scrubber for AI agents. Sanitizes untrusted text and blocks detected wallet-draining commands before context ingestion.

SSRF-safe public GET with status, headers, and size-capped JSON/text body for agent tool use

Classify a caller-supplied operation claim against caller-supplied evidence under deterministic rules and return a machine-readable handoff receipt. Use this when an orchestrator needs a consistent completion state before accepting an agent's work. VERIFIED_COMPLETE means the submitted evidence supports the submitted claim; QSL does not independently observe the underlying real-world action.

Verify factual claims and check whether cited sources support them. Fact-check statistics, dates and source attributions in agent research and reports against web evidence, preserving reporting periods, units and population groups. Returns a verdict, explanation and cited sources; /v2/verify links evidence to source IDs. 0.20 USDC via x402 v2 on Base. Model-assessed; no signup or API key.

Use when an agent needs the best live price for a stock across venues before trading. For 1-10 tickers: perp mark and mid on trade.xyz (Hyperliquid), Aster, Lighter and OKX, xStocks and Ondo token prices per share, last US close, the session reference price, gap % vs it, cross-venue spread, best bid/ask and freshness. Unpaid calls get 402; bad or unknown tickers get 400/404 before any charge. Erro

x402-native person discovery for autonomous agents. Given search criteria (title, seniority, company, industry, or location) instead of a known name, returns a list of matching people with title, seniority, department, decision-maker status, LinkedIn URL, and location. Structured JSON. Paid per request via x402.

The Bouncer's log for ERC-8004 agents on Base, up to 100 entries an answer, newest first: fraud flags (self-hire, payment loops, shared funders, bursts, review farms) with severity and onchain receipts (tx hashes, wallets, the measured metric and its threshold), and reviewer cohorts (wallets that started reviewing within hours of each other, with their funding window and first-review txs). Call it

Use when an agent is about to pay an x402 endpoint, seller domain or wallet it has not used before. Returns ok_to_pay, caution, dont_pay or no_payment_needed with reasons: live payment terms, its Coinbase Bazaar listing and paying wallets in 30 days, whether the live pay-to address matches the listing, real-USDC check, price vs. similar tools, and seller domain age. Read-only: never pays or signs.

Is this transaction, token approval or signature safe to sign? Check it before your AI agent or wallet signs it: pre-sign verdict (green/orange/red + reason codes) for EVM and Solana transactions, token approvals and Permit/Permit2/EIP-3009/Seaport signatures. Screens the spender or recipient wallet for OFAC sanctions and scam, phishing or theft reports, and flags unlimited allowances, plain-walle

Use when an agent needs fresh identifiers — UUID v4/v7, ULID, nanoid, secure passwords, or memorable passphrases. Cryptographically random via WebCrypto, generated locally, no upstream. Pass type and count (1-100). Use permitted: build it into your products and services, paid or free. Raw redistribution of the feed itself is prohibited.

[Payment decision: spend policy + seller reputation risk check, one verdict] Right before an AI agent pays: POST {payment, policy, seller} -> ONE verdict pay / dont_pay / ask_human. Runs the spend-policy check (limits, rolling cap, categories, allow/deny, hours, assets/chains, approval threshold) and the x402 seller reputation check (buyers, wash share, age, concentration), with reasons and one si

Financial metrics per fiscal period for a company by ticker or CIK, computed from SEC filings: valuation (P/E, P/B, EV/EBITDA), margins, returns, liquidity, leverage, growth and per-share figures. Period annual, quarterly or ttm; filter by report_period or filing_date. Use when an agent compares valuation or profitability across periods or companies. Non-Professional use; Professional use needs a

Liquidity pool data for agents scanning yield opportunities: TVL, APY, volume, and IL risk for the top 500 DeFi pools, filterable by protocol/chain/token, refreshed every 15 minutes. Category: data.

Is this x402 endpoint safe to pay? Check it before your AI agent pays an endpoint it has not used before. Returns a go / caution / no_go verdict, the recommended payment option (network and amount) and the reasons (payment would fail, payout address on the OFAC sanctions list, over your budget, price above advertised, not HTTPS, unknown token). Never pays the endpoint. Cached 10 min. Bulk: 1000 cr

YouTube transcript and video search API for AI agents: retrieve timestamped transcript segments by video ID or find videos by keyword. Returns JSON transcripts with text, start and duration, or video search results for content research and summarization. Provided by FastAF YouTube Transcript & Metadata. Price: $0.001000 USDC per successful call.

Anchor an AI agent decision or audit record to Base mainnet. Submit {schema, payload_hash (SHA-256 hex, 64 chars), metadata}; service queues into a Merkle batch and returns a receipt with pending_anchor status immediately. Full Merkle proof and on-chain anchor_tx available via free GET /v1/notarize/:receipt_id after batch anchors (within batch_threshold receipts or 120s). Include Idempotency-Key h

Smart contract security check before you approve or pay an address: reads the bytecode, behind proxies too, for privileged functions (mint, pause, blacklist, upgrade), dangerous opcodes and protocol type, and checks whether a single key holds those powers. Adds OFAC SDN screening, who funded the address, sybil-filtered ERC-8004 reputation for agents and a 0-100 score with proceed / hold / block. S

Screen an EVM wallet or contract address BEFORE paying it. Deterministic, keyless compliance / safety gate: checks the address against the OFAC sanctioned crypto-address list and a curated taint set (mixer, hack, scam) and returns a single BLOCK / REVIEW / ALLOW recommendation plus the matched sources, so an AI agent can decide whether to send funds. No API keys, no LLM, one x402 USDC micropayment

Live feed of every active rail disruption across the Japanese (JP) network in one call — delays and suspensions with bilingual summaries, refreshed continuously from ODPT. Built for MONITORING agents: poll on an interval (e.g. every 5 minutes) to maintain a nationwide incident watchlist without querying stations one by one.

A stateful, multi-chain spending session for AI agents. One session enforces a hard USDC budget cap across every tool call, with a verifiable receipt and running ledger for each payment — not a one-shot budget check, but persistent spend governance with a full audit trail. The gateway enforces the cap for the paying wallet on Stacks and Base. USDC on Base (standard x402) or Stellar (via the AgentP

Curated reading list on QE for tutoring agents, quiz generators, and reading apps — 168 public-domain books (e.g. A Study of Recent Earthquakes, The chemistry, properties and tests of precious stones, Field book of common rocks and minerals : $b for identifying the rocks and minerals of the United States and interpreting their origins and meanings). Full-text retrieval with drill-down routes for c

One call, many sources: search Hacker News, GitHub, Stack Overflow, and Wikipedia by keyword and get back a unified, ranked result set — titles, URLs, scores/stars, and snippets — across all of them at once. The general research/search read agents make to ground answers, find code, surface discussion, and gather context. No API keys, no per-source accounts, free upstreams. Pass `sources` to narrow

URL safety preflight: screen a link before an agent opens it — syntax, host/IP, patterns, redirects, optional threat intelligence.

Is this company or domain legit? Business and domain verification (KYB) for vendor screening, supplier checks and scam or fraud review: domain age and registrar from RDAP, DNS (A/MX/NS), email security (SPF/DMARC) and HTTPS. Returns a verdict and a 0-100 trust score with evidence URLs, and flags newly registered or thin-infrastructure domains, a common scam signal. Use it before your agent pays an

SECONDED checks EVM transactions, permits, tokens, agent registrations and suspicious messages before you sign or act. Two rival-lab models must agree; NOT VERIFIED means pause or ask a human. Choose a product and submit its exact input. A paid answer is a signed check, not a guarantee.

Multi-platform tokenized-equity premium/discount intelligence for AI trading agents — 24/7 on-chain tokenized stocks (xStocks NVDAx, Ondo Global Markets NVDAon, Dinari dShares where readable) vs the underlying share: premium/discount in bps per platform, cross-platform spread when 2+ platforms trade the same name, US session state, on-chain liquidity, and a mean-reversion read. Loud about platform

Paid book chapter (x402): Automate the Boring Stuff with Python: What Is Programming. Unit of a larger practical-skills library; free catalog at /library/discovery.json. Book slug: automatetheboringstuffwithpythonnew. Response format: Markdown (.md) for agent parsing. USDC on Base via Coinbase CDP facilitator. Free teasers and discovery.json require no payment.

On-chain Forensics API for AI agents: token verdict cards and settle-safe pre-trade screens ($0.02), ERC-20 burn/mint and wallet forensics ($0.05, Base/Ethereum/XRPL), and $0.25 due-diligence memos. Bare URL = $0.02 directory tier; append ?endpoint=<name> to call a data endpoint

Use this before an AI agent pays an x402 endpoint: give the endpoint URL, we read its own 402 quote, pull every payTo (all networks) and screen each against OFAC, UN, EU and UK lists. Signed per-payee and overall verdict: deny on an OFAC match, review on a UN/EU/UK match or an unavailable list, clean otherwise. $0.01 in USDC on Base or Solana. A bad, unreachable or non-x402 URL returns 400, 422 or

DCL Trust Oracle. Call only when the text must include the substring AI and must not include "I cannot be held responsible" or "no guarantees". POST JSON {"response","agent_id"}. COMMIT means those checks passed and the action may proceed. NO_COMMIT means do not execute the checked action. Not a tool-call guardrail, prompt-injection detector, or jailbreak screen; use /evaluate/fast for the phrase