Tool results (MCP servers, APIs, search, file reads) are untrusted content the agent re-reads. Detects a result that gives orders to the agent, strays from what the tool should return or asks to send data (tool poisoning). Returns allow, review or block.

Batch EU EORI validation via official EC EOS (up to 100 per request)

Company name search — find Legal Entity Identifiers (LEIs), legal names, jurisdictions and registration status. Get up to ten matches per page to identify counterparties, distinguish similarly named companies and enrich company reference data.

Peru policy interest rate — latest monthly reference-rate observation in annual percent. Track monetary policy trends and benchmark borrowing costs with the reference month.

Hallucination check for RAG: is a claim supported by the source text? Returns supported, unsupported, contradicted or uncertain with probabilities, so an agent can verify an answer before showing it.

Agent security guardrail. Judges the pair (untrusted content + proposed action) before an AI agent acts: detects tool calls that obey hidden prompt injection, leave the user's goal, destroy data or send data or money elsewhere. Adjustable sensitivity and optional multi-model consensus. Returns allow

For coding agents that run shell commands: approve, review or block. Text rules first (root deletion, download-and-run, key exfiltration, reverse shells, SQL destruction, persistence, privilege escalation), then a fake terminal that simulates the command without ever running it (decoy files, pipe an

Output guardrail for customer-support chatbots. Blocks unauthorized promises, invented facts, leaked instructions or customer data, off-topic answers and bad tone before the reply is sent. Returns send, human_review or block.

Base token check before you buy: owner, mint, upgrade, pause and blacklist powers read on-chain. Facts only, not advice.

Before an agent clicks "I accept": flags risky clauses in terms of service or contracts (auto-renewal, data resale, waiver of court or class action, liability cap, unilateral changes, exit penalties, IP grab). Flags, does not give legal advice. Returns accept, review_clauses or do_not_accept.

Inbound lead qualification and routing for sales agents: spam, job application, vendor pitch, existing customer, or a real project with urgency. Hard rules first (disposable email, tests), returns a routing lane.

Before an agent installs an MCP server, skill or tool: is the description poisoned (hidden orders to the agent), does it ask for more permissions than its purpose, does it imitate a known tool? Code checks (lookalike names, broad permissions, hidden text) then a reading of the descriptions. Judges t

Invoice-fraud guard for AI agents that pay bills. Detects supplier bank-detail change fraud (BEC) before payment: checks the IBAN and sender domain against your supplier record (lookalike domains included), then reads pressure and off-channel requests. Returns ok, verify_callback or freeze_payment.

Federal campaign finance profile: who a candidate is, what they raised and spent, and what is left in the bank.

Sentiment from -1 to +1 with confidence, mixed-opinion flag and a separate score per aspect you name (price, support, delivery). Abstains when the text is ambiguous.

US motor carrier safety verdict by USDOT: authority, out-of-service, rating, crashes. Unverified raw FMCSA data, ~24h old.

College verdict: cost, size, admission rate, completion, median earnings and debt for a named US institution.

Who launched this Solana coin, and did their other coins graduate or collapse? Serial launcher check.

PII detector that finds what regex misses (names, addresses, ID numbers, health data, credentials) and locates what patterns can verify (emails, valid IBANs, Luhn-checked cards, phones). Does not rewrite text.

Sort text into one of your own labels, with a confidence and a one-line reason. You supply the labels.