{"uid":"cap_QHIMf8Xp6fWrq3x0-wjUC","slug":"agent-run-review-3c95de49","name":"Agent Run Review","description":"817 pay-per-call APIs for AI agents. Every endpoint is POST /<name> and is payable in USDC over x402 on Base and Solana — no signup, no API key, just pay-per-call. GET https://x402.agentutility.ai/ returns the same catalog as machine-readable JSON.","url":"https://x402.agentutility.ai/agent-run-review?utm_source=zero.xyz","method":"POST","headers":{},"bodySchema":{"type":"object","$schema":"https://json-schema.org/draft/2020-12/schema","required":["input"],"properties":{"input":{"type":"object","required":["type","method","bodyType","body"],"properties":{"body":{"required":["trace"],"properties":{"trace":{"type":["array","string"],"description":"The agent execution trace to review: a JSON array of steps/tool calls, or a plain-text log. Required. Serialized form capped at 40,000 chars."},"models":{"type":"array","items":{"type":"string"},"description":"Optional list of model names to price the run's estimated tokens against, e.g. ['gpt-4o-mini','claude-sonnet-4.5']. Defaults to a standard model spread."}}},"type":{"type":"string","const":"http"},"method":{"enum":["POST"],"type":"string"},"bodyType":{"enum":["json","form-data","text"],"type":"string"}},"additionalProperties":false},"output":{"type":"object","required":["type"],"properties":{"type":{"type":"string"},"example":{"type":"object","properties":{"review":{"type":"object","properties":{"unknowns":{"type":"array"},"confidence":{"type":"string"},"step_summary":{"type":"array","items":{"type":"object","properties":{"step":{"type":"integer"},"action":{"type":"string"},"outcome":{"type":"string"}}}},"failure_points":{"type":"array","items":{"type":"string"}},"root_cause_hypothesis":{"type":"string"}}},"degraded":{"type":"boolean"},"components":{"type":"array","items":{"type":"object","properties":{"ms":{"type":"integer"},"ok":{"type":"boolean"},"name":{"type":"string"}}}},"composed_of":{"type":"array","items":{"type":"string"}},"trace_chars":{"type":"integer"},"cost_estimate":{"type":"object","properties":{"estimates":{"type":"array","items":{"type":"object","properties":{"model":{"type":"string"},"total_cost_usd":{"type":"number"}}}},"input_tokens":{"type":"integer"},"cheapest_known":{"type":"string"}}},"generated_evals":{"type":"object","properties":{"cases":{"type":"array","items":{"type":"object","properties":{"input":{"type":"string"},"rubric":{"type":"string"},"expected":{"type":"string"}}}},"returned_n":{"type":"integer"},"review_required":{"type":"boolean"}}}}}}}}},"responseSchema":{"type":"json","example":{"review":{"unknowns":[],"confidence":"medium","step_summary":[{"step":1,"action":"search_web","outcome":"returned 3 results"}],"failure_points":["step 2: parse_result threw TypeError on result[0]"],"root_cause_hypothesis":"search_web returned a result shape parse_result did not expect."},"degraded":false,"components":[{"ms":1400,"ok":true,"name":"agent-trace-brief"}],"composed_of":["agent-trace-brief","llm-cost-estimate","eval-case-generate"],"trace_chars":140,"cost_estimate":{"estimates":[{"model":"gpt-4o-mini","total_cost_usd":0.00002}],"input_tokens":120,"cheapest_known":"gpt-4o-mini"},"generated_evals":{"cases":[{"input":"x","rubric":"check y","expected":"y"}],"returned_n":8,"review_required":true}}},"example":null,"exampleRequest":null,"tags":["x402"],"displayCostAmount":"0.06","displayCostAsset":"USDC","priceDynamic":false,"priceHint":null,"priceStatus":"priced","priceSource":"registry","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":"$0.06/call","primary":{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.06","per":"call","confidence":"exact"},"accepted":[{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.06","per":"call","confidence":"exact"}]},"paymentMethods":[{"uid":"pm_-hQuOKfCOp5m3yUN38HhK","protocol":"x402","methodType":"crypto","chain":"base","mode":"charge","costAmount":"0.06","costPer":"request","priority":0,"asset":null,"unit":"request","depositMicros":null,"planRef":null}],"brandName":null,"brandSlug":null,"brandBaseUrl":null,"brandDocsUrl":null,"whatItDoes":"Analyzes an AI agent execution trace to identify failure points, estimate token costs, and generate evaluation test cases","exampleAgentPrompt":"Can you review this agent execution trace and tell me where it failed, why it failed, and what it would have cost on gpt-4o-mini and claude-sonnet-4-5? Here's the trace: [{'step':1,'tool':'search_web','result':'ok'},{'step':2,'tool':'parse_result','error':'TypeError'}]","exampleUseCases":[{"title":"Post-mortem on failed agent run","prompt":"My agent crashed midway through a task — here's the full execution trace as a JSON array. Can you review it, identify what step failed and why, and give me a root cause hypothesis?"},{"title":"Token cost breakdown by model","prompt":"I just ran an agent and want to know what it would have cost across different models. Here's the trace — can you estimate the token cost for gpt-4o-mini and claude-sonnet-4-5?"},{"title":"Auto-generate evals from agent trace","prompt":"I have a working agent trace I want to turn into regression tests. Can you analyze this trace and generate evaluation cases with inputs, rubrics, and expected outputs so I can catch future regressions?"}],"resultDescription":"Returns a structured review object with: a step-by-step summary of what the agent did, identified failure points with descriptions, a root cause hypothesis, a confidence rating (low/medium/high), unknown/ambiguous steps, token cost estimates across one or more LLM models, and auto-generated evaluation test cases derived from the trace. Also includes latency and health data for internal sub-components used to produce the review.","failureModes":["Trace too long — serialized trace exceeds 40,000 character limit, request rejected","Malformed trace — trace is neither a valid JSON array nor parseable plain text, review quality degrades","Unknown model names — unrecognized model identifiers in the models list, those models omitted from cost estimate","Empty trace — no steps detected, returns minimal review with low confidence","Degraded mode — one or more internal sub-components (e.g. eval generator) fail, response still returned but degraded flag set to true"],"whenToPreferThis":"Use this endpoint when you have a completed or failed AI agent execution trace and need structured post-mortem analysis — specifically failure diagnosis, root cause identification, LLM cost estimation, and auto-generated eval cases. Prefer this over manual log inspection when the trace is complex or multi-step. It is particularly valuable for agent developers building observability pipelines, debugging flaky agent behavior, or bootstrapping eval suites from real runs.","instructions":null,"reviewSummary":null,"reviewSummaryHighlights":null,"reviewSummaryConcerns":null,"reviewSummaryGeneratedAt":null,"activationCount":0,"lastUsedAt":null,"lastSuccessfullyRanAt":null,"lastHealthCheckAt":"2026-11-21T18:31:26.102Z","isFirstParty":false,"canonicalSlug":"agent-run-review-3c95de49"}