{"uid":"cap_gusjXg6nlh3YXPTxlVks1","slug":"verdict-l402-untrusted-text-guardrail-73622e57","name":"Verdict L402 – Untrusted Text Guardrail","description":"Pre-flight guardrail for untrusted text before an LLM agent reads it: prompt injection, data exfiltration requests, secrets and API keys, personal data, toxicity. A la carte: pick the checks (price drops with them), the sensitivity (strict, balanced, lenient) and an optional consensus of independent","url":"https://verdict.kechlab.com/v1/guard?utm_source=zero.xyz","method":"POST","headers":{},"bodySchema":{"type":"object","properties":{"text":{"type":"string","description":"the untrusted text to examine"},"checks":{"type":"array","items":{"type":"string"},"description":"(optional) list among injection, exfiltration, pii, secrets, toxic: a single check costs half"},"consensus":{"type":"boolean","description":"(optional) true: three independent models, disagreement = review, double price"},"sensitivity":{"enum":["strict","balanced","lenient"],"type":"string","description":"(optional) strict, balanced (default) or lenient"}}},"responseSchema":null,"example":null,"exampleRequest":null,"tags":["x402"],"displayCostAmount":"0.0174","displayCostAsset":"USDC","priceDynamic":false,"priceHint":null,"priceStatus":"priced","priceSource":"probe","requiresHandshake":false,"reviewCount":0,"rating":{"score":"0.00","successRate":"0.00","reviews":0,"stars":null,"state":"unrated"},"availabilityStatus":"unknown","priceObserved":null,"sessionDeposit":null,"pricing":{"kind":"static","summary":"$0.0174/call","primary":{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.0174","per":"call","confidence":"exact"},"accepted":[{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.0174","per":"call","confidence":"exact"}]},"paymentMethods":[{"uid":"pm_ylOzOYfypbwmAgHKXRif-","protocol":"x402","methodType":"crypto","chain":"base","mode":"charge","costAmount":"0.0174","costPer":"request","priority":0,"asset":"0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913","unit":"request","depositMicros":null,"planRef":null}],"brandName":null,"brandSlug":null,"brandBaseUrl":null,"brandDocsUrl":null,"whatItDoes":"Scans untrusted text for prompt injection, data exfiltration attempts, secrets/API keys, PII, and toxicity before an LLM agent processes it","exampleAgentPrompt":"Before you process this user-submitted message, scan it with balanced sensitivity for prompt injection and data exfiltration attempts — block it if anything looks suspicious.","exampleUseCases":[{"title":"RAG pipeline injection shield","prompt":"Before feeding this web-scraped document into my RAG pipeline, scan it for prompt injection and data exfiltration attempts using strict sensitivity — I want three independent models to agree before it passes."},{"title":"User chat input PII and secrets filter","prompt":"Check this incoming user message for PII and accidentally included secrets or API keys before my agent reads it — use balanced sensitivity and just those two checks to keep the cost down."},{"title":"Toxic content pre-screening for agent inbox","prompt":"I'm getting untrusted messages from external users. Screen this one for toxicity and prompt injection using lenient sensitivity before my assistant replies to it."}],"resultDescription":"Returns a verdict indicating whether the text is safe to process, which specific threat categories were triggered (injection, exfiltration, pii, secrets, toxic), and if consensus mode is used, whether the models disagreed (flagging it for human review). Allows downstream agent to decide to pass, block, or escalate.","failureModes":["Text too long may exceed limits and return an error","Invalid check type names return a validation error","Consensus mode with high disagreement returns a 'review' verdict rather than a clear pass/block","Misconfigured sensitivity level defaults to balanced or returns a 400 error","Network timeout on the $0.0176 payment flow if L402 auth is not handled correctly"],"whenToPreferThis":"Choose this endpoint when your agent ingests untrusted text from external sources (user input, web scraping, external APIs, email) and you need a pre-flight safety check before the LLM processes it. It is specifically designed for agentic pipelines and covers the full threat surface (injection, exfiltration, PII, secrets, toxicity) in a single call. Prefer it over generic content moderation APIs when prompt injection and secret leakage are primary concerns. The à la carte check selection makes it cost-efficient when you only need a subset of protections.","instructions":null,"reviewSummary":null,"reviewSummaryHighlights":null,"reviewSummaryConcerns":null,"reviewSummaryGeneratedAt":null,"activationCount":0,"lastUsedAt":null,"lastSuccessfullyRanAt":null,"lastHealthCheckAt":"2026-10-21T07:20:16.351Z","isFirstParty":false,"canonicalSlug":"verdict-l402-untrusted-text-guardrail-73622e57"}