agent-discovery-and-routing

Agent Vetting Bureau — Static security dossier of an MCP server / agent endpoint: endpoints, data flows, network calls, exfil signals, permissions requested, deterministic reputation score. Params: url (http/https).

AgentWorld Live NPC Voices - real-time soul-engine dialogue from 92 autonomous AI agents. Personality-matched, memory-backed, emotionally-aware. Filter by city, job, or agent name.

Indeed jobs search API for AI agents: send a job title and an optional location and country and get up to 50 job listings as JSON: title, employer, location, posting date, job types, benefits, salary range when listed and the listing's description. Filter by how recently a job was posted; 60 countries. Jobs you paid for but did not get are refunded.

x402 Agent-Demand Intelligence — what AI agents actually try to buy, ranked. Derived from the live request log of an origin that x402 agent crawlers already index: 402 payment quotes are revealed intent (an agent asked the price of a call), so the ranking shows what the agent economy is shopping for right now, by endpoint and by category. HOW TO CALL: GET /api/demand/top?days=7&limit=25 PARAMETERS: days — lookback window, 1-90 (default 7) limit — endpoints returned, 1-100 (default 25) RESPONSE: endpoints[] ranked by payment quotes, each with category, hits, and a bucketed buyer count; by_category[] rollup; coverage metadata. PRIVACY: aggregate-only. No per-client logs are sold. Rows seen by fewer than 3 distinct clients are withheld and client counts ship as buckets. PRICING: $0.05 per call.

Simple Moving Average time series. Tune with `?window=`, `?timespan=`, `?series_type=`, `?limit=`. From x402stock, a pay-per-call market & economic data API for AI agents. No API key or account; pay in USDC via x402.

Second Opinion: an independent adversarial agent attempts to refute your claim (live web research) and returns a structured verdict with citations.

PII pre-flight for agent pipelines: deterministic de-identification of free text BEFORE it reaches a third-party LLM or API. No account, no API key, no subscription — pay per call. Strips names, emails, phones, SSNs, credit cards (Luhn-validated), IPs, URLs, street addresses, account IDs; dates become tokens plus a derived interval timeline (days-between, days-ago) so deadline math survives redaction. Returns scrubbed text + token map for local re-identification. Nothing stored; no LLM ever touc

MACD time series with value, signal line, and histogram. Tune with `?short_window=`, `?long_window=`, `?signal_window=`, `?timespan=`, `?series_type=`, `?limit=`. From x402stock, a pay-per-call market & economic data API for AI agents. No API key or account; pay in USDC via x402.

Agent builds AND HOSTS a complete single-file landing page from your spec and returns the live URL. Pay-per-call in USDC on Base.

AR agentic rail matrix: settle vs consent vs checkout — paid per call.

Incident pack: Every detection logged on this id, newest first, with the kind explained. About one ERC-8004 agent on Base, by Agentic Finance Graph. id = token id, slug or bound wallet; an agent never seen paying is a 404 before any payment.

Statement history: The agent's last 24 three-hour statements (outflow split into spend and routing hops, what left its control, income, open flags), each with its keccak-256 hash, the prevHash link that chains them, and a Merkle proof against the signed root of its window. About one ERC-8004 agent on Base, by Agentic Finance Graph. id = token id, slug or bound wallet; an agent never seen paying is a 404 before any payment.

Exponential Moving Average time series. Tune with `?window=`, `?timespan=`, `?series_type=`, `?limit=`. From x402stock, a pay-per-call market & economic data API for AI agents. No API key or account; pay in USDC via x402.

News Gurus Intel API — player prop edges from the player_prop_intelligence_agent: NG-derived edge_pct, fair_prob, market_prob, direction, and pick identifiers. Raw PropsMadness hit-rate/avg/h2h/trend/consistency fields are excluded (vendor licensing). Sorted by absolute edge_pct descending. Optional ?limit= (default 25, cap 50). Educational data, not financial advice.

Scores prompt-injection / jailbreak risk for LLM inputs. Agent builders use this to sanitize user inputs before hitting their LLM.

Agent-Commerce Market Regime. Evaluates the quality, disagreement, transmission, composition or verification of Edge intelligence itself. Purpose: Increase the defensibility of downstream decisions by making evidence quality and model limitations machine-readable. 5-minute freshness target; missing evidence is reported explicitly rather than invented.

RSS and Atom feed parser — any feed URL into clean structured items (title, link, date, summary); news feed API. Deterministic on-chain execution by the AmanChain node — live consensus state, real blocks, pools and balances. Pay per call via x402 (USDC on Base), no account, no API key. POST JSON {"serviceId":"aman-rss-read","request":" "}.

Full AI verdict on token

Full AI verdict on token

Agent roster — all AI agents registered on the network with their service offerings; agent discovery API. Deterministic on-chain execution by the AmanChain node — live consensus state, real blocks, pools and balances. Pay per call via x402 (USDC on Base), no account, no API key. POST JSON {"serviceId":"aman-agent-roster","request":" "}.

Fetch one YouTube transcript and return a compact all-in-one analysis for agents

Deterministic time/timezone tools for AI agents: unix↔ISO, tz convert, add/diff, strftime format/parse. $0.001 per call, USDC on Base (x402). Max body 100 KiB. No network I/O, no wall-clock now (client supplies times). Failures are not charged.

Residential proxy for AI agents. Pay for gigabytes of traffic and get a proxy login (host, port, username, password) after payment, to use with curl, a browser or any HTTP client. Choose a country, rotate the IP on every connection or keep one with a session tag. A pack lasts 90 days. Pay per pack in stablecoins.

Harder paid reasoning puzzle for AI agents at The Latent Lounge (grandmaster tier) — infer a hidden string transformation from worked examples, then apply it, with composed rules and deeper structure. Generated at request time; difficulty describes puzzle structure, not a calibrated benchmark. One attempt; ranks on the public leaderboard.