agent-discovery-and-routing

Agent Delivery Order Check: Agent Delivery Order Check checks delivered identifiers against expected order from bounded caller-supplied values without an external provider. Call Agent Delivery Order Check before paying for or accepting an x402 or agent delivery assembled from caller-supplied challenge and result metadata. Returns contract-specific checks, normalized evidence, mismatch reasons, and an explicit acceptance or rejection status for Agent Delivery Order Check as versioned determinist…

Compact chat on Granite 4.0 H Micro (@cf/ibm-granite/granite-4.0-h-micro). Lowest floor cost for simple agent loops and classification-style prompts.

Validates an Agent Card document against the A2A convention and returns findings by severity: error (the card is unusable as published), warning (works, but a consuming agent will guess), info (recommended field absent). Checked: the required fields name, description, url, version, capabilities and skills; that url is absolute and https; that every skill carries id, name, description and tags and that skill ids are unique; that declared securitySchemes use a known type and that every entry in…

Takes the steps of one agent run and returns a scored report. It counts identical tool calls (same tool, same arguments) as loops, finds the longest run of consecutive failures, marks calls whose result was empty or null as wasted, and attributes cost per step and per tool so the expensive part is visible. Cost is taken from your figures when supplied, otherwise computed from token counts against the model price table.

Drug safety profile for AI agents: given a US drug name (brand or generic), returns FDA label highlights (boxed warning, contraindications, key interactions), pregnancy/pediatric/geriatric flags, pharmacologic class, top adverse-event signals, recent recalls, current shortage status, and marketing/discontinued status. RxNorm-normalized name resolution with a confidence score. Data: openFDA (CC0). Not for clinical use.

Runtime intelligence action verification with signed ALLOW, CONSTRAIN, or DENY decision.

Client Asset / File Request agent pack (mf-tpl-034)

QC Fail / Rework (Internal) agent pack (mf-tpl-026)

Unforgeable operating history for any AI agent attesting through Rubric Protocol: record count, first-seen date, continuity days, and recent activity, computed from HCS-anchored attestations that cannot be backdated.

Unforgeable operating history for any AI agent attesting through Rubric: record count, first-seen date, continuity days and recent activity — computed from HCS-anchored attestations that cannot be backdated.

Attested x402 payment history for any buyer wallet: settlement count, first-seen, last-seen, total spent, services used, distinct sellers observed - computed from an append-only settlement ledger and anchored on Hedera. Counterparty intelligence that is evidence, not opinion: unlike behavioral scores, these records cannot be revised.

Retrieves recovery assistance or troubleshooting guidance for base-level agent issues.

Live AI-citation check: runs fresh probes against major AI assistants and reports whether a domain is cited for its vertical's buying questions — citation rate per provider, list position, competitors named instead, and whether the domain's agent-card was discovered. Fresh probes per call (not cached). $0.50 per check. Free companion diagnosis: /api/x402/deep-scan.

Counterparty trust verification for agents: before transacting with a domain, verify its trust artifacts — DNSSEC + DNS-AID records, agent-card, live MCP endpoint, published JWKS, OASF record, x402/commerce declarations. Returns a 0-100 trust score with per-artifact evidence. $0.25 per lookup.

Silver tier — 100-word premium AI experience + leaderboard entry

Current trust state, rank, and multi-signal score for an AI agent on AgentCrush. Updated every 4 hours.

Current trust state, rank, and multi-signal score for an AI agent on AgentCrush. Updated every 4 hours.

Wertet robots.txt fuer 23 benannte KI-Agenten aus, mit korrekter Gruppenauswahl und der Regel, dass die laengste passende Anweisung gewinnt. Prueft zusaetzlich X-Robots-Tag, Meta-Angaben, noai, Bezahlschranken-Hinweise, Lizenzverweise, Nutzungsbedingungen und den TDM-Vorbehalt unter /.well-known/tdmrep.json.

Python code execution as a service for agents: send code, get stdout/stderr/exit code. Isolated, hard limits. $0.01 USDC x402 Base.

Base address -> agent-ready on-chain profile: EOA or contract, ETH and major-token holdings, token or NFT metadata, EIP-1967 proxy detection, activity, and a derived profile label. Uses public Base mainnet RPC with no account or API key. Useful before an agent sends funds, swaps, or calls a contract.

State of Agent Commerce report data — full structured JSON, one query

AI Act Article 50 disclosure - the privacy-first GET path: pass only the SHA-256 of your AI output (content_sha256), never the text itself. Get back the signed provenance record, the human-readable disclosure in DE and EN, and the obligations that apply to the deployer. Your content never leaves your system. Use POST with 'content' if you want us to hash it for you.

Agent Card Validator (A2A): given an Agent Card URL or JSON, deterministically validates it against the A2A spec — required fields, skill schemas, endpoint reachability, security schemes, provider transparency — and returns a compatibility_score, schema_errors and warnings. The trust layer for the agent economy. Deterministic, Ed25519-signed. 0.01 USDC via x402 on Base. SYNTHORA.

Gives an agent without a reachable server a place to receive callbacks. create returns a public URL that accepts any POST, GET or PUT from a third party; poll returns the events that arrived since a cursor you carry forward, with headers, parsed body and arrival time. The inbox expires after at most seven days and holds at most 500 events, oldest dropped first. Reading requires the secret returned at creation.