tag:agentutility.ai,2026:updatesagentutility — updatesDaily notes on x402, the agent economy, and the agentutility endpoint portfolio. By x.com/agentutility_ai.2026-10-07T12:00:00Zagentutility-sitehttps://agentutility.ai/icon.pngagentutilityhttps://agentutility.aitag:agentutility.ai,2026-10-07:/updates/exchange-ratesexchange-rates: convert currency with a rate date you can record2026-10-07T12:00:00Z2026-10-07T12:00:00ZRequest current or historical currency rates with exchange-rates, convert a specific amount, and keep the returned rate date beside the value in your billing or reporting workflow.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-10-06:/updates/bridge-ratesbridge-rates: compare output and costs before signing a bridge transfer2026-10-06T12:00:00Z2026-10-06T12:00:00ZInspect a cross-chain route with bridge-rates: supply every required input, compare expected output and costs, then check timing and the transaction details you'll need before signing.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-10-05:/updates/cve-lookupcve-lookup: get severity and exploit references from a CVE ID2026-10-05T12:00:00Z2026-10-05T12:00:00ZSend a CVE identifier to cve-lookup for severity data and affected product identifiers, then use weakness classes and linked exploit evidence to build a triage record.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-10-04:/updates/sku-normalizesku-normalize: turn product titles into catalog matching keys2026-10-04T12:00:00Z2026-10-04T12:00:00ZTurn messy product titles into structured attributes and a canonical catalog key with sku-normalize, then use the confidence value to decide which records need review before matching.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-10-03:/updates/warranty-parsewarranty-parse: extract coverage limits and claim steps2026-10-03T12:00:00Z2026-10-03T12:00:00ZTurn warranty wording into coverage fields your agent can check before a purchase or claim, with duration in months, stated exclusions, claim instructions, and explicit unknowns.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-10-02:/updates/return-policy-extractreturn-policy-extract: read return windows and missing refund terms2026-10-02T12:00:00Z2026-10-02T12:00:00ZSend merchant policy text to return-policy-extract and read structured refund terms, with null values for unstated details so your agent can ask for missing information before acting.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-10-01:/updates/browser-render-repeat-checksBrowser render for repeat page checks2026-10-01T12:00:00Z2026-10-01T12:00:00ZA runnable pattern for checking JavaScript-rendered pages again and comparing the returned page state.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-30:/updates/honeypot-check-update-2026-09-30honeypot-check now names the pre-swap sell-block use case2026-09-30T12:00:00Z2026-09-30T12:00:00ZThe honeypot-check description now names token honeypot checks and pre-swap sell-block checks explicitly, giving agents clearer routing language while the price stays at 0.05 USDC per call.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-29:/updates/product-searchproduct-search: find retailer buy links with detected prices2026-09-29T12:00:00Z2026-09-29T12:00:00ZSend a product query and an optional retailer domain to product-search, then use ranked buy links and detected prices to build a shopping shortlist your agent can check.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-27:/updates/product-describeproduct-describe: write listing copy from your product facts2026-09-27T12:00:00Z2026-09-27T12:00:00ZTurn supplied product specifications into listing descriptions and selling points with product-describe, with request and response examples for agents preparing retail copy.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-27:/updates/news-searchRun the same news query on a schedule2026-09-27T12:00:00Z2026-09-27T12:00:00Znews-search now documents the repeat-query workflow for tracking fresh coverage by URL and timestamp.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-26:/updates/pdf-to-markdownpdf-to-markdown: choose the document structure your next step needs2026-09-26T12:00:00Z2026-09-26T12:00:00ZSend a public PDF URL to pdf-to-markdown and choose the document structure your workflow needs, from Markdown for an LLM prompt to JSON blocks for page-aware extraction.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-25:/updates/timezone-lookuptimezone-lookup: get local time from coordinates or a timezone2026-09-25T12:00:00Z2026-09-25T12:00:00ZResolve coordinates or an IANA timezone into current local time, UTC offset, and daylight-saving status so your agent can check contact hours and prepare schedules with the right timezone.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-24:/updates/form-fill-planform-fill-plan: review browser actions before filling a form2026-09-24T12:00:00Z2026-09-24T12:00:00ZPrepare a browser form plan from static HTML and a stated goal, using known values to review proposed actions, catch unresolved fields, and hold submission until the caller has checked the result.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-22:/updates/dom-change-diffdom-change-diff: check stale selectors before a browser rerun2026-09-22T12:00:00Z2026-09-22T12:00:00ZCompare earlier and current HTML with watched selectors to find changed elements and stale targets, then check suggested replacements before your agent reruns a browser task.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-20:/updates/crypto-news-update-2026-09-20crypto-news adds aggregator wording for agent discovery2026-09-20T12:00:00Z2026-09-20T12:00:00ZThe crypto-news description now names its aggregator use case, giving routers clearer wording for headline searches while keeping the $0.005 USDC price and request fields unchanged.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-19:/updates/browser-flow-digestbrowser-flow-digest: turn browser traces into steps an agent can reuse2026-09-19T12:00:00Z2026-09-19T12:00:00ZSend a browser action log or text trace to browser-flow-digest and get ordered steps with preconditions, failure hints, and reuse notes that help an agent prepare its next attempt.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-17:/updates/tool-call-difftool-call-diff: catch changed arguments and missing calls between agent runs2026-09-17T12:00:00Z2026-09-17T12:00:00ZCompare baseline and changed agent runs with tool-call-diff to spot added or missing calls, shifted positions, and exact argument-path changes before accepting a regression.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-16:/updates/mcp-tool-risk-scoremcp-tool-risk-score: review tool permissions before granting access2026-09-16T12:00:00Z2026-09-16T12:00:00ZScreen an MCP tool manifest with deployment context, then use per-tool reasons and risk scores to decide which permissions need review before your agent gets access.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-15:/updates/eval-case-generateeval-case-generate: turn an agent task into cases for human review2026-09-15T12:00:00Z2026-09-15T12:00:00ZDraft candidate eval cases from an agent task and optional examples, then use the review_required signal to keep unapproved expectations out of your regression test suite.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-14:/updates/agent-trace-briefagent-trace-brief: find the failure point in an agent run2026-09-14T12:00:00Z2026-09-14T12:00:00ZSend an agent trace as step objects or a text log, add an optional focus hint, and get an ordered report that separates observed failures from explanations the trace can't support.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-13:/updates/match-scorematch-score: put possible duplicate records into a review queue2026-09-13T12:00:00Z2026-09-13T12:00:00ZCompare two field values for $0.003 USDC with match-score, then use its similarity score and normalized output to route possible duplicate records into a review queue.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-10:/updates/web-search-empty-fallbackweb-search now flags an empty fallback as unavailable2026-09-10T12:00:00Z2026-09-10T12:00:00Zweb-search returns a structured dependency response when the primary search fails and the public-news fallback has no articles.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-09:/updates/web-search-repeat-proofweb-search refreshes its repeat-research proof2026-09-09T12:00:00Z2026-09-09T12:00:00ZThe web-search recipe now carries the latest aggregate repeat-buyer snapshot for its saved source-list comparison workflow.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-09:/updates/recent-tweet-search-monitoring-looprecent-tweet-search now describes a practical monitoring loop2026-09-09T12:00:00Z2026-09-09T12:00:00ZThe listing and MCP tool now explain how to save a query, compare successive responses, and handle a degraded public-index result honestly.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-08:/updates/search-result-scopeSearch results now enforce site and purchase constraints2026-09-08T12:00:00Z2026-09-08T12:00:00ZSite searches filter off-domain URLs, and product search removes discussion pages from purchase candidates.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-07:/updates/match-key-phonematch-key-phone: turn phone formatting variants into matching contact keys2026-09-07T12:00:00Z2026-09-07T12:00:00ZUse match-key-phone to compare differently formatted contact numbers, read its E.164 output and validity flag, and handle region inputs before grouping possible duplicates.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-06:/updates/match-key-namematch-key-name: match Bob Smith with Robert Smith2026-09-06T12:00:00Z2026-09-06T12:00:00ZSee how match-key-name handles titles and suffixes, folds accents, and maps nicknames to shared keys, with exact response examples for connecting person records across sources.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-05:/updates/match-key-emailmatch-key-email groups Gmail aliases under one deduplication key2026-09-05T12:00:00Z2026-09-05T12:00:00ZSee how match-key-email turns Gmail dots and plus tags into one deduplication key, then add email-deliverability-check when domain risk or mail readiness affects the next action.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-04:/updates/match-key-companymatch-key-company: turn business-name variants into one dedup key2026-09-04T12:00:00Z2026-09-04T12:00:00ZSee how punctuation, legal suffixes, abbreviations, and misspellings change normalized company names and match keys before you deduplicate records for 0.005 USDC per call.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-03:/updates/news-apiCall recent public news through the news-api route2026-09-03T12:00:00Z2026-09-03T12:00:00Znews-api gives x402 agents the existing news-search contract under the exact route name they ask for, at $0.005 per call.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-02:/updates/match-key-addressmatch-key-address: join US address variants with one key2026-09-02T12:00:00Z2026-09-02T12:00:00ZFor $0.005 USDC over x402, match-key-address gives equivalent US address spellings one join key; match-score ranks close records whose keys still differ.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-09-01:/updates/address-parseaddress-parse: turn one US mailing address into eight fields2026-09-01T12:00:00Z2026-09-01T12:00:00ZSend a freeform US mailing address to address-parse, get each component as a named field for $0.005 USDC, and learn when your workflow should call address-geocode next.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-31:/updates/image-generate-fal-failoverimage-generate adds fal.ai failover2026-08-31T12:00:00Z2026-08-31T12:00:00ZPaid image requests can move to fal.ai Z-Image Turbo when Venice returns a definite provider failure.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-31:/updates/image-generate-controlled-next-passimage-generate now names the controlled next-pass job2026-08-31T12:00:00Z2026-08-31T12:00:00ZThe image-generate listing and Synthforge MCP tool now lead with the approved-brief workflow: keep the setup, then make one deliberate next PNG.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-31:/updates/holder-behavioral-profilingHolder data that separates exits from wallet moves2026-08-31T12:00:00Z2026-08-31T12:00:00Z>-agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-31:/updates/holder-analysis-reliabilityHolder analysis reliability improvements2026-08-31T12:00:00Z2026-08-31T12:00:00Zagentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-28:/updates/token-momentumCheck whether a token is moving now2026-08-28T12:00:00Z2026-08-28T12:00:00ZThe $0.003 token-momentum route compares short-window price, volume, and transaction pressure on the token's most liquid DEX pair.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-28:/updates/new-pairsPoll newly created DEX pools2026-08-28T12:00:00Z2026-08-28T12:00:00ZThe $0.003 new-pairs route returns recent Base, Solana, or Ink pools with liquidity, age, and market telemetry.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-28:/updates/mint-scoutScan NFT mints on Robinhood Chain and Ink2026-08-28T12:00:00Z2026-08-28T12:00:00ZThe $0.05 mint-scout route finds fresh NFT collections or scores one address with mint, holder, creator, and liveness data.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-28:/updates/holder-handsMeasure NFT holder churn2026-08-28T12:00:00Z2026-08-28T12:00:00ZThe $0.05 holder-hands route samples holders and recent NFT transfers to measure concentration, flips, and wash signals.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-28:/updates/deployer-profileCheck a contract creator's prior deploys2026-08-28T12:00:00Z2026-08-28T12:00:00ZThe $0.005 deployer-profile route resolves a creator and returns bounded public Blockscout history for Base, Ink, or Robinhood Chain.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-27:/updates/buy-link-searchFind a product and return the buy link2026-08-27T12:00:00Z2026-08-27T12:00:00ZThe $0.02 buy-link-search route returns ranked shopping results with retailer links and detected prices.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-26:/updates/ssl-cert-monitoring-for-agentsCheck TLS certificates before a scraping agent spends on a crawl2026-08-26T12:00:00Z2026-08-26T12:00:00ZUse WebProbe's $0.03 ssl-cert call to inspect hostname coverage, issuer, validity dates, and expiry risk before a scraping agent commits budget to a target.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-24:/updates/data-blending-policyWhen agentutility blends sources and how responses show it2026-08-24T12:00:00Z2026-08-24T12:00:00ZAgentutility combines upstream sources only under a declared method, with paid x402 responses exposing field-level provenance, timestamps, source selection, and conflicts.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-23:/updates/audit-readiness-for-paid-apisAudit-ready x402 calls need receipts and idempotency keys2026-08-23T12:00:00Z2026-08-23T12:00:00ZEach paid agent call should carry an idempotency key and return a receipt tying the request to its Base mainnet payment, price, result, and timestamp.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-22:/updates/web-search-page-text-boundaryweb-search now states its result-only boundary2026-08-22T12:00:00Z2026-08-22T12:00:00ZThe web-search family now labels include_text as compatibility-only and directs page-text work to scrape-to-json.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-22:/updates/the-cost-of-running-a-clusterWhat a $0.01 x402 call has to pay for2026-08-22T12:00:00Z2026-08-22T12:00:00ZA $0.01 x402 call must cover Workers, KV, external API usage, Base settlement, and monitoring. Here’s a per-endpoint margin model for setting a price that can stay online.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-21:/updates/weather-endpoint-coverageLocale weather covers coordinates worldwide2026-08-21T12:00:00Z2026-08-21T12:00:00ZThe Locale weather routes combine US NWS forecasts with a global model blend, returning current conditions and up to seven forecast days for coordinates worldwide.agentutilityhttps://agentutility.aitag:agentutility.ai,2026-08-20:/updates/agent-side-retry-logicRetry paid x402 calls without losing budget control2026-08-20T12:00:00Z2026-08-20T12:00:00ZSeparate x402 payment handshakes from server retries, stop repeated 402 loops, cap USDC spend, and add exponential backoff without turning a transient 5xx into a budget leak.agentutilityhttps://agentutility.ai