{"uid":"cap_o74hIcvGCX42QM1Z0m15p","slug":"agent-alexweb-space-deep-verify-4163dfd6","name":"agent.alexweb.space Deep Verify","description":"Four-pass independent and adversarial verification for high-confidence agent QA.","url":"https://agent.alexweb.space/v1/deep-verify","method":"POST","headers":{},"bodySchema":{"type":"object","properties":{"task":{"type":"string","description":"Original task or requirement"},"candidate":{"type":"string","description":"Candidate output to verify"},"constraints":{"type":"array","items":{"type":"string"},"description":"Optional hard constraints"}}},"responseSchema":null,"example":null,"exampleRequest":null,"tags":["x402"],"displayCostAmount":"0.2","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":"$0.2/call","primary":{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.2","per":"call","confidence":"exact"},"accepted":[{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.2","per":"call","confidence":"exact"}]},"paymentMethods":[{"uid":"pm_j2QfcHpJW_wlSmkbqVPfT","protocol":"x402","methodType":"crypto","chain":"base","mode":"charge","costAmount":"0.2","costPer":"request","priority":0,"asset":"0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913","unit":"request","depositMicros":null,"planRef":null}],"brandName":null,"brandSlug":null,"brandBaseUrl":null,"brandDocsUrl":null,"whatItDoes":"Runs a four-pass independent and adversarial verification pipeline to assess agent-generated outputs against a task and optional hard constraints.","exampleAgentPrompt":"Run a deep four-pass adversarial verification on this candidate answer — the original task was 'summarize the Q3 earnings report in 3 bullet points', the candidate output is '• Revenue up 12% YoY • Net income $4.2B • Guidance raised for Q4', and the hard constraints are ['no speculation', 'cite only numbers from the report'].","exampleUseCases":[{"title":"Pre-deployment agent output QA","prompt":"Before we publish this AI-generated product description, run a thorough four-pass adversarial check — the task was 'write a 100-word product description for our noise-cancelling headphones', here's the candidate text: 'Experience silence like never before…', and the constraints are ['under 110 words', 'no unverifiable claims', 'mention battery life']."},{"title":"Automated code review gate","prompt":"I need high-confidence verification on this code summary my agent produced — the original task was 'explain what this Python function does', the candidate explanation is 'This function sorts a list in descending order using a quicksort algorithm', and the constraint is ['must be technically accurate', 'reference time complexity']."},{"title":"Legal draft compliance check","prompt":"Can you do a deep adversarial verify on this contract clause draft? The task was 'draft a liability limitation clause for a SaaS agreement', the candidate clause is 'In no event shall Provider be liable for indirect damages…', and the hard constraints are ['must comply with California law', 'cap must be stated explicitly']."}],"resultDescription":"A structured multi-pass verification report indicating whether the candidate output satisfies the original task, including adversarial critiques from each independent pass, identification of constraint violations, and an overall confidence or pass/fail determination.","failureModes":["Missing required 'task' or 'candidate' fields returns a 400 error","Overly long or malformed candidate text may be rejected or truncated","If constraints are contradictory, the verification may flag all passes as failing without clear resolution","Network timeouts on very large inputs given the multi-pass latency profile","Payment failure via x402 protocol blocks the request before processing"],"whenToPreferThis":"Choose this endpoint when you need maximum confidence in an agent's output — especially before high-stakes actions like publishing, deploying, or filing. It is a four-pass adversarial process (vs. the sibling two-pass fast verify), making it slower but significantly more thorough. Prefer it when hard constraints must be enforced, when errors are costly, or when a single-pass review is insufficient for your risk tolerance.","instructions":null,"reviewSummary":null,"reviewSummaryHighlights":null,"reviewSummaryConcerns":null,"reviewSummaryGeneratedAt":null,"activationCount":0,"lastUsedAt":null,"lastSuccessfullyRanAt":null,"lastHealthCheckAt":"2026-09-14T12:37:12.686Z","isFirstParty":false}