@agentutility/mcp-compose

One call, several of our endpoints, one settlement.

Composite endpoints that run two to five of our own capabilities in parallel and return one merged result. A company dossier is enrichment plus whois plus TLS plus tech stack in a single paid call. You skip the orchestration, the retries, and four extra settlements.

Pricing: pay-per-call in USDC on Base. No subscriptions, no API keys. See per-tool prices below.

Install — Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "agentutility-compose": {
      "command": "npx",
      "args": ["-y", "@agentutility/mcp-compose"],
      "env": { "X402_PRIVATE_KEY": "0xYOUR_PRIVATE_KEY_HEX" }
    }
  }
}

Restart Claude Desktop. 132 tools appear in the tool palette.

Install — Cursor

Add to .cursor/mcp.json:

{
  "mcpServers": {
    "agentutility-compose": {
      "command": "npx",
      "args": ["-y", "@agentutility/mcp-compose"],
      "env": { "X402_PRIVATE_KEY": "0x..." }
    }
  }
}

Funding

Send any amount of USDC on Base mainnet to the address derived from your X402_PRIVATE_KEY. The MCP server uses it to pay for tool calls automatically.

USDC on Base contract: 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913

Tools (132)

| Tool | Description | |---|---| | address-intel | (0.03 USDC/call) Address intelligence API for a street address: parse, geocode, timezone, and building permit context in one call. Composite: one call runs address-parse + address-geocode + timezone-lookup + building-permits and merges the results into one address intel record. Same composite backend as address-intel-pack. Use it as address intel, address intelligence, location enrichment, property context, or real-estate/logistics address enrichment for agents. | | address-intel-pack | (0.03 USDC/call) Enriches a street address in one call: parse, geocode, timezone, and building permits. Composite: address-parse + address-geocode run in parallel, then timezone-lookup on the geocoded coordinates and building-permits when the address resolves to a supported open-data city (NYC, SF, Chicago, LA, Seattle, Austin). Merged output {parsed, location, timezone, permits}. Open-data sources only (OSM Nominatim, TimeAPI, city Socrata portals). Use it as an address intelligence API or location enrichment and property context for real-estate and logistics agents. | | address-intelligence | (0.03 USDC/call) Enriches an address with parsed components, geocoding, timezone, and available city building-permit context. Composite: one call runs address-parse + address-geocode + timezone-lookup + building-permits and returns parsed, location, timezone, and permits fields. Same composite backend as address-intel-pack. Use it as an address intelligence API, property intelligence endpoint, or location enrichment API. | | agent-input-guard | (0.04 USDC/call) One guardrail call an agent makes before trusting untrusted inbound text — a scraped webpage, an email, a tool result, or a user message. Composite: one call runs prompt-injection-detect + moderate-content + ai-content-detector in parallel, returning one allow/review/block verdict instead of three separate settlements. prompt-injection-detect is the required leg — a failure there fails the call; moderate-content and ai-content-detector degrade the response instead of failing it on a rate limit or model hiccup. Verdict blocks on a confirmed injection or a severe moderation category, reviews on a suspicious signal or borderline category, and allows otherwise. Use it as a prompt injection filter, agent input firewall, untrusted content screen, or pre-LLM guardrail API. | | agent-prelaunch-audit | (0.08 USDC/call) Ship-readiness check before you push an AI app live: agent due diligence and prelaunch audit in one call instead of six. Inspects deploy config, secret exposure, prompt-injection surface, dependency risk, and database migration risk, then rolls the result into one readiness view. Composite: one call runs deploy-config-risk + secrets-exposure-check + prompt-injection-surface + dep-risk-summary + db-migration-risk + production-readiness-score in parallel into {deploy_config, secrets, prompt_injection, dependencies, db_migration, readiness_score}. Send a repo ('owner/name') for a live scan, or any mix of config, deps, prompt, and sql; each maps to its matching check, so partial input still returns a partial verdict. Advisory and heuristic only, it flags patterns and never executes code. Use it as an ai app security check, a pre-deploy audit gate, or a ship-readiness score. | | agent-run-review | (0.06 USDC/call) Agent run review / trace cost evals / post-run analysis / agent debugging in one call: hand it a raw agent execution trace and get a readable brief of what happened, a cost estimate for the run, and a first draft of regression eval cases to catch the same failure next time. Composite: one call runs agent-trace-brief + llm-cost-estimate + eval-case-generate in parallel and merges them into {review, cost_estimate, generated_evals}. agent-trace-brief turns the raw steps/tool-calls into an ordered summary plus a root-cause hypothesis for any failure; llm-cost-estimate prices the run's tokens across a model list; eval-case-generate drafts candidate {input, expected, rubric} cases from the inferred task so the workflow gets regression coverage. Use it as an agent debugging API, a post-run trace reviewer, or a step toward turning production incidents into eval sets. | | article-brief | (0.04 USDC/call) Analyzes a news article from its URL into a summary, named entities, and sentiment in one call. Returns scraped text stats, a bullet summary, entities (people, organizations, locations, dates), and overall sentiment with score: the standard agent reading pipeline collapsed into a single request. Ideal for news monitoring, research agents, content triage, brand tracking. Composite: one call runs scrape first, then summarize-text + extract-entities + sentiment in parallel over the text. Use it as a URL-to-summary article analyzer or one-call article intelligence. | | audio-brief | (0.07 USDC/call) Transcribes and digests any audio URL in one call: transcript, summary, and sentiment. Composite: one call runs audio-transcribe + summarize-text + sentiment. Whisper large v3 transcription, then summary and sentiment analysis run in parallel over the transcript. Returns transcript, detected language, duration, summary, sentiment scores, and per-component telemetry. Use it for podcast summaries, call recording digests, or speech-to-text plus TLDR plus emotion. | | authentic-sentiment | (0.06 USDC/call) Authentic sentiment API: send a topic, brand, handle, or phrase and get back sentiment from real humans, not bots or AI astroturf. Composite: one call runs x-search for recent posts, then ai-content-detector and x-account-authenticity per post, then sentiment-analysis, excluding posts that read as AI-generated (probability_ai_generated at or above 0.65) or from low-authenticity authors (score under 40) before rolling up sentiment. Post search is required; the per-post AI-detection, authenticity, and sentiment legs are degradable, so a rate limit narrows the rollup instead of failing the call. Returns the filtered authentic_sentiment read plus a raw_sentiment_all_posts figure and exclusion counts. Use it as an authentic sentiment API, real-vs-bot sentiment tool, bot-filtered brand sentiment monitor, or AI-noise-filtered social sentiment tool. | | brand-availability-sweep | (0.06 USDC/call) Brand availability sweep API for a name availability check before you commit to a startup, product, or project name: trademark signal, domain availability, X/Twitter handle availability, brandable alternative names, and a brandability score in one call. Composite: one call runs brand-clearance + domain-availability + x-handle-availability + domain-suggest + brand-name-score in parallel (the handle and alternative-name legs are optional, degrading if X credentials are missing or the naming model is slow) into {clearance, primary_domain_availability, handle, alternatives, brandability}. Trademark data is informational only, not legal clearance. Use it as a brand availability API, name availability check, or trademark-and-domain search. | | brand-presence-report | (0.10 USDC/call) Screens a brand or startup name in one call: domain availability, social handle, clearance signals, and lookalike risk. Composite: one call runs domain-availability + social-handle-check + brand-clearance + homoglyph-check in parallel and merges them into {domains, social_handle, clearance, lookalike_risk}. Domain availability is checked per requested TLD (default com, ai, io, dev; max 6), the X handle is probed read-only, brand-clearance aggregates USPTO/domain/Wikipedia/HN signals, and homoglyph-check scores Unicode lookalike risk. Screening data only; not legal, trademark, or brand advice. Use it as a brand presence report API, brand name check, or startup name screening bundle for naming and launch agents. | | browser-workflow-repair | (0.06 USDC/call) Browser workflow repair / self-healing automation / dom drift fix / flow memory in one call for agents whose CSS selectors broke after a page redeploy. Composite: one call runs dom-change-diff + form-fill-plan + browser-flow-digest + website-screenshot in parallel, merged into {dom_diff, form_fill_plan, flow_digest, screenshot}. dom-change-diff compares before/after HTML and suggests stable replacement selectors for anything that moved; form-fill-plan drafts a fresh fill plan against current markup and goal; browser-flow-digest compresses a prior trajectory into reusable workflow memory (only runs when supplied); website-screenshot captures the live page when a url is given. Use it as a self-healing browser automation API, a DOM drift repair tool, or a workflow-memory generator. | | caption-translate-pack | (0.07 USDC/call) Generates translated subtitles from any video URL in one call. Composite: one call runs video-to-subtitles + translate. SRT subtitles are generated via Whisper v3, then the subtitle text is translated into the target language (100+ languages, structure preserved). Returns original subtitles, translated subtitles, and per-component telemetry. Use it for localized video captions, an SRT generator plus translation, or subtitle localization. | | claim-corroborate | (0.05 USDC/call) Fact check a claim and see exactly who agrees and who disagrees. Composite: one call runs web-search + wikipedia-search + x-search + arxiv-search + citation-verify + sentiment in parallel, then re-verifies the top hits against their source pages and rolls up a corroboration count. Send a claim and get back a per-source stance (supports, contradicts, or unclear) with the snippet it was judged on, plus an overall corroboration score. Sources without a conclusive citation-verify read fall back to a keyword-overlap heuristic instead of guessing. Use it as a fact check API, claim verification tool, source cross-reference for research agents, or a misinformation screen before an agent repeats a claim downstream. | | clean-text-pack | (0.03 USDC/call) Sanitizes raw text in one call: Unicode normalization, homoglyph audit, PII detection, and PII redaction. Returns the NFC-normalized string, a script/homoglyph/hidden-character audit (Cyrillic lookalikes, zero-width chars, RTL overrides), every PII match with risk level, and a fully redacted copy with [EMAIL]/[PHONE]/[SSN]-style masks. One paid call replaces four sanitization steps: ideal for log scrubbing, dataset prep, LLM input hygiene, phishing triage. Composite: one call runs text-normalize + unicode-normalize + detect-pii + pii-redact in parallel over the same text. Use it as a text sanitizer, GDPR-safe ingest pipeline, or input scrubber. | | commodity-brief | (0.05 USDC/call) Market brief for a commodity: recency-scoped web search headlines on price/supply/demand plus a structured research brief with overview, key facts, and recent news with source citations, in one call. Composite: one call runs search as the required leg and research-brief in parallel as an optional leg, so an LLM/scrape hiccup degrades the response instead of failing the call. Use it as a commodity market brief or price/supply/demand digest for agents tracking oil, metals, grains, or other commodities. | | company-dossier | (0.05 USDC/call) Builds a company research dossier in one call: firmographic profile, domain intelligence, WHOIS registration, TLS, and tech stack. Composite: one call runs company-enrich + domain-enrich + whois + ssl-cert-info + tech-stack-detect in parallel and merges them into {profile, domain, registration, security, tech_stack}. Accepts a domain or a bare company name (name inputs are resolved via a best-guess domain probe; unresolvable names return the profile only with the rest marked skipped). Public-data only; no private enrichment databases. Use it as a company background lookup or account research for sales and B2B agents. | | company-intel-pack | (0.08 USDC/call) Company research bundle for B2B enrichment agents: firmographic profile, domain intelligence, WHOIS registration, TLS posture, tech stack, email authentication (DMARC/SPF/DKIM), and recent news in one call. Composite: one call runs company-enrich + domain-enrich + whois + ssl-cert-info + tech-stack-detect + dmarc-check + news-search in parallel and merges them into {profile, domain, registration, security, tech_stack, email_auth, news}. Accepts a company name or a domain (name inputs resolve via a best-guess domain probe; unresolvable names return the profile only with the rest marked skipped). tech-stack-detect and news-search are optional legs that degrade instead of failing the pack. Seven lookups bought separately add up to $0.105; this pack is $0.08. Public-data only; no private enrichment databases. | | company-verify-pack | (0.04 USDC/call) Checks that a company exists and its public signals are consistent, in one call: profile, registrar, domain age, and TLS. Composite: one call runs company-enrich + registrar-lookup + domain-age-check + ssl-cert-info in parallel and merges them into {profile, registrar, age, tls}. Public-data existence and consistency signals only; this is data verification from open sources, NOT compliance, KYB, KYC, or identity verification. Use it as a company verification data API, vendor data consistency check, or domain-backed company signals for procurement and marketplace agents. | | competitor-snapshot | (0.06 USDC/call) Snapshots a company for competitive intelligence in one call: profile, recent news, and news sentiment. Composite: one call runs company-enrich + web-search + brand-sentiment-analysis; the profile and a week-recency ' news' search run in parallel, then sentiment is scored over the top result snippets when the search returns hits. Merged output {profile, news, sentiment}. Data snapshot only; no analysis, recommendations, or investment advice. Use it as a company news monitor or market-research data feed. | | contact-golden-record | (0.06 USDC/call) Turn 2-5 messy candidate contact records into one deduplicated golden record with per-field provenance. Composite: one call runs people-enrich + company-enrich + match-key-email + match-key-name + match-key-company + match-score in parallel, then clusters the records that describe the same person and merges them, picking the highest-confidence value per field and keeping track of which input it came from. Send name/email/company/phone variants for the same contact and get back one merged record, a match-confidence score, and a list of field-level conflicts it had to resolve. Use it as an entity resolution API, contact dedupe tool, CRM record merge, or identity-resolution step before writing a lead into a database. | | contact-normalize-pack | (0.03 USDC/call) Normalizes a contact record in one call, generating stable match keys for email, name, phone, and address plus parsed US address components. Composite: one call runs match-key-email + match-key-name + match-key-phone + match-key-address + address-parse in parallel; only the constituents whose input field is present are invoked (at least one of email/name/phone/address required). Deterministic and instant — no external lookups, no data retention. Use it as a contact normalization API, contact dedupe bundle, match key generator, record linkage pack, or CRM hygiene tool. | | content-authenticity-report | (0.06 USDC/call) Content authenticity report: send a URL (or raw text) and find out whether a page is AI-generated slop before your agent cites, summarizes, or acts on it. Distinct from content-trust-check (moderation + originality on text you already hold): this starts from a live URL and also reads the page's lead images. Composite: one call runs scrape (required, also sources up to two lead images) then, in parallel, ai-content-detector on the text (required when there is enough text) and ai-image-detector on each lead image (optional, degradable). Raw text input skips extraction and images. Returns a per-text read, a per-image read, and one authenticity_score (0-1, 1 = authentic) blending text 70% and images 30%. Both detectors are calibrated heuristics, not forensic-level. Use it as a content authenticity checker, AI slop filter for web pages, source quality screen, or a pre-citation check. | | content-quality-pack | (0.04 USDC/call) Runs the standard pre-publish content checks on text in one call: AI-detection, PII scan, moderation, and sentiment. Returns an AI-generated probability with verdict, every PII match with risk level, an allow/review/block moderation verdict, and overall sentiment. Ideal for UGC pipelines, comment gating, guest-post screening, agent output review. Composite: one call runs ai-content-detector + detect-pii + moderate-content + sentiment in parallel over the same text. Use it as a pre-publish content check or UGC safety gate. | | content-trust-check | (0.05 USDC/call) Trust screen for a block of text in one call: is it AI-generated, is it safe (content moderation), and is it a near-duplicate of a reference page, plus a deterministic 0-100 trust score rolling all three up. Composite: one call runs ai-content-detector + moderate-content + content-simhash in parallel; when reference_url is supplied, it also scrapes that page and simhashes it to compute a Hamming-distance near-duplicate check against the input text. At least one of the three core legs must succeed; the others degrade the response instead of failing the call. The trust_score is computed in plain code from the legs' outputs, not an LLM call. Use it as an AI content detector, plagiarism/originality check, content moderation gate, or pre-publish trust screen for agents vetting text before it goes out. | | contract-terms-brief | (0.05 USDC/call) Pull SLA terms, named entities, PII flags, and a plain-language summary out of a contract or policy document in one call. Built for contract review, SLA extraction, agreement summary, and terms extraction workflows where an agent needs structured contract terms brief data before a human reads the full document. This is extraction only, not legal advice: it reports what the text states and does not interpret enforceability. Composite: one call runs sla-extract + extract-entities + detect-pii + summarize-text in parallel and merges results into named sections. Sends contract text (alias contract_text); returns sla_terms, entities, pii, and summary. Use it as a contract review assistant, SLA extraction API, or agreement summary tool. | | contract-trust-pack | (0.04 USDC/call) Gathers smart-contract due-diligence data in one call: source verification, honeypot simulation, and LP lock check. Composite: one call runs contract-source-verified + token-honeypot + lp-lock-check in parallel and merges Etherscan-verified source status (name, compiler, ABI, proxy flag), buy/sell tax simulation results, and liquidity-locker holdings into one structured report with per-component telemetry. Long Solidity sources are truncated in the merged payload. Data-only aggregation; no advice. Use it as a smart-contract verification bundle or EVM contract due-diligence data pack. | | country-risk-monitor | (0.08 USDC/call) Re-runnable country monitor: recency-scoped web search headlines, any live Polymarket prediction markets referencing the country (data-only, no trading), and a structured research brief, in one call. Composite: one call runs search as the required leg, then polymarket-markets and research-brief in parallel as optional legs, so a quiet prediction market or an LLM hiccup degrades the response instead of failing the call. Informational monitoring only — no investment, travel-safety, or immigration advice framing anywhere in the output. Use it as a recurring country-situation monitor for agents tracking a country's political and economic backdrop. | | course-study-kit | (0.05 USDC/call) Turn a course syllabus into a study kit in one call: a parsed outline, a sequenced reading/study plan, flashcards, and a bullet summary. Built for course prep, study kit, syllabus to plan, and learning workflow use cases where a student or agent needs a syllabus breakdown without doing four separate lookups. This generates a study outline and aids from the syllabus text provided; it makes no credentialing, credit, or accreditation claims. Composite: one call runs reading-plan-generate + paper-to-flashcards + summarize-text in parallel, after syllabus-parse extracts the course name. Sends syllabus_text plus optional hours and level; returns outline, plan, flashcards, and summary. Use it as a syllabus breakdown tool, study plan generator, or course prep assistant. | | cron-pack | (0.02 USDC/call) Parses, explains, and forecasts a cron expression in one call. Composite: one call runs cron-parse + cron-explain + cron-next in parallel over the same 5/6/7-field cron expression or @daily/@hourly/@weekly/@monthly/@yearly macro. Returns validation + frequency bucket, a plain-English explanation, and the next N firing times as ISO 8601 + epoch seconds + human-readable strings in your timezone, with regular-cadence detection. Includes composed_of + per-component telemetry; partial failure degrades instead of failing. Use it as a crontab validator, plain-English schedule explainer, or 'when does this cron run' forecast. | | cross-border-ops-brief | (0.05 USDC/call) Cross-border operating brief for a country: the FX rate, country profile, visa requirements, upcoming public holidays, and current local time an agent needs before transacting or scheduling internationally. Composite: one call runs country-info first to resolve the country, then exchange-rates, visa-requirements, holiday-lookup, and timezone-lookup in parallel, merged into one brief. Give it a country name or ISO code plus optional home_currency and nationality; get back fx, visa, holidays, and local_time sections. Reference data only, not money transmission or immigration advice. Use it as an international operations API, business-travel brief, or country-risk reference for agents planning cross-border work. | | cross-lingual-topic-monitor | (0.06 USDC/call) Monitor a topic across non-English sources in one call: search the open web and X, detect what language each hit is actually in, translate it into your target language, and score sentiment on the translated text. Composite: one call runs web-search, x-search, detect-language, translate, and sentiment in parallel and per-hit and merges them into a language-tagged, sentiment-scored feed. Send a topic, a target_lang, and an optional day/week/month window; get back each hit's source, url, original text, detected language, translated text, and sentiment, plus a language_mix breakdown. Use it as a multilingual monitoring API, a foreign news tracker, or a global topic-tracking and cross-lingual sentiment tool. | | crypto-news-brief | (0.03 USDC/call) Daily crypto news digest for a topic: today's headlines plus a structured research brief with overview, key facts, and recent news with source citations, in one call. Composite: one call runs search (day-recency headlines) as the required leg and research-brief in parallel as an optional leg, so a single scrape or LLM hiccup on the brief degrades the response instead of failing the call since the raw headlines already stand on their own. Use it as a daily crypto news roundup or topic digest for agents tracking a coin, protocol, or crypto theme. | | crypto-signal-scan | (0.05 USDC/call) Structured, informational signal brief for a token or crypto topic: X/Twitter sentiment, recent web coverage, and (when a fiat quote currency is supplied) a current exchange-rate reference, in one call. Composite: one call runs twitter-sentiment + search in parallel (both required together — the call only fails if both fail) plus exchange-rates as an optional third leg, since most crypto tickers aren't ECB-quoted fiat pairs. Data snapshot only: no trade signals, price targets, or investment recommendations. Use it as a crypto sentiment and news snapshot for agents tracking a token or topic across social and web sources. | | cyber-exposure-brief | (0.08 USDC/call) External exposure summary for a domain: DNS records, the hosting network/ASN behind the resolved IP, TLS certificate posture, and a web search for public breach/exposure mentions, in one call. Composite: one call runs dns-lookup as a required first stage (it supplies the IP the ip-asn leg needs), then ip-asn + ssl-cert-info + search in parallel as a fully optional second stage, so any one probe timing out degrades the response instead of failing the call. Distinct from domain-dossier (whois/dns/ssl/domain-age/domain-risk/dmarc, registration framing): network/threat-surface framing, no repeat of domain-dossier's whois, age, risk, or DMARC legs. Data-collection snapshot only — not a penetration test, vulnerability scan, or security certification. Use it as an external exposure brief for security teams scouting a domain's attack surface. | | defi-protocol-dossier | (0.04 USDC/call) Profiles a DeFi protocol's TVL and yield pools in one call using DeFiLlama data. Composite: one call runs defi-llama-protocol + defi-yield-pools in parallel, merging the protocol's canonical profile (name, category, audits, current TVL, per-chain TVL breakdown, 24h/7d/30d change) with its yield-bearing pools filtered to the same project slug (APY, TVL, exposure, IL-risk classification). Pool list is optional and degrades without failing the call. Data-only aggregation; no advice. Use it as a DeFiLlama protocol profile bundle or an Aave / Lido / Uniswap TVL and APY data pack. | | dependency-provenance | (0.06 USDC/call) Dependency risk and supply chain trust for one package in a single call: registry adoption stats, a package risk score, source repo health and star velocity, auto-routed to the right ecosystem. Composite: one call runs npm-package-stats or pypi-package-stats or crates-io-stats, package-risk-npm or pypi-package-risk, github-repo-health, and github-stars-history in parallel, ecosystem-routed from the package name or an explicit ecosystem input. Send a package name and an optional ecosystem or repo override; get back stats, a risk score, repo health and stars trend, merged into one trust view. Use it as a package trust API, an sbom enrichment tool, or a supply-chain risk check before you install a dependency. | | destination-arrival-pack | (0.05 USDC/call) On-arrival brief for a city or airport: this is the arrival pack, destination brief, and local conditions check a travel agent pulls right after landing. Composite: one call runs weather-current-city + air-quality + timezone-lookup + sunrise-sunset + exchange-rates in parallel, after resolving the destination. Returns airport metadata (when destination is an IATA/ICAO code), current weather, air quality index and pollutants, local time and UTC offset, sunrise/sunset/twilight times, and a home-to-local currency exchange rate. Send destination as a city name or a 3/4-letter IATA/ICAO code, plus an optional home_currency. Use it as an arrival pack API, destination brief, or city-on-arrival check for travel and relocation agents. | | doc-brief | (0.05 USDC/call) Digests a public PDF in one call: layout-aware markdown, an executive summary, and named entities. Markdown preserves headings, tables, and lists; entities cover people, organizations, dates, and monetary amounts. Handles digital and scanned PDFs up to 30 pages. Ideal for report triage, contract review prep, research ingestion, due-diligence agents. Composite: one call runs pdf-to-markdown first, then summarize-text + extract-entities in parallel over the markdown. Use it as a PDF summary + entities bundle, PDF intelligence, report digest API, or contract skim. | | document-extract-suite | (0.07 USDC/call) Turn a PDF URL into structured data in one call: this pdf extraction suite pulls markdown text, tables, named entities, PII flags, and a summary without five separate requests. Composite: one call runs pdf-to-markdown first, then fans pdf-extract-tables, extract-entities, detect-pii, and summarize-text in parallel over the extracted text. Send { url: '' } and get back { markdown, tables, entities, pii_flags, summary }. Built for document intelligence api workflows, pdf to structured data pipelines, and any document parse job feeding an LLM or a database. Only a hosted PDF URL is supported — there is no raw file-upload path. If entity/PII/summary legs fail, markdown and tables still return; the response is marked degraded rather than failed. | | domain-dossier | (0.04 USDC/call) Builds a full domain report in one call: WHOIS, DNS, TLS, age, risk, and DMARC. Composite: one call runs whois + dns-lookup + ssl-cert-info + domain-age-check + domain-risk-check + dmarc-check in parallel and merges them into {registration, dns, tls, age, risk, email_auth}. One request replaces six separate lookups for security, fraud, and due-diligence agents. Use it as a domain intelligence API, phishing triage bundle, or brand-protection domain workup. | | domain-threat-report | (0.08 USDC/call) Domain threat report API for a security investigation on any domain: is this a phishing domain, a lookalike squat, or a legitimate registration? Covers WHOIS/RDAP registration age, DNS records, TLS certificate posture, DMARC/SPF/DKIM email authentication, certificate-transparency subdomain enumeration, IP/ASN hosting attribution, and homoglyph lookalike detection on the domain string. Composite: one call runs whois-lookup + dns-lookup + ssl-cert-info + dmarc-check + subdomain-enum + homoglyph-check + ip-asn in parallel (ip-asn needs a resolved A record and is skipped otherwise) into {registration, dns, tls, email_auth, subdomains, ip_intelligence, lookalike_check}. Use it as a phishing check API, domain security audit, or brand-impersonation detector. | | email-trust-report | (0.03 USDC/call) Vets an email address in one call: deliverability, disposable check, DMARC posture, and a dedupe key. Composite: one call runs email-deliverability-check + disposable-email-check + dmarc-check + match-key-email in parallel and merges them into {deliverability, disposable, email_auth, match_key}. dmarc-check audits the email's domain; match-key-email supplies a normalized record-linkage key. DNS + public blocklist data only; no mailbox probing. Use it as an email verification bundle or email risk data for signup vetting and CRM hygiene agents. | | entity-knowledge-card | (0.05 USDC/call) Entity knowledge card API for entity resolution and entity enrich: send a name (person, place, organization, or concept) and get back a structured knowledge graph profile plus a readable summary in one call. Built for agents that need to disambiguate a name or build a knowledge card without stitching a Wikidata lookup to a Wikipedia fetch by hand. Composite: one call runs wikidata-resolve + wikipedia-search + wikipedia-lookup + company-enrich in parallel. wikidata-resolve is required and returns the QID, type claims, and key claims; wikipedia-search and wikipedia-lookup add a readable summary and disambiguation candidates; company-enrich is a best-effort org profile probe, only meaningful when the entity is a company, and degrades silently otherwise. Use it as an entity knowledge card API, entity resolution tool, or knowledge graph lookup. | | esg-company-scan | (0.08 USDC/call) Summary of a company's publicly reported ESG posture: a web search scoped to sustainability/emissions/labor/governance coverage plus a structured research brief with overview, key facts, and recent news with source citations, in one call. Composite: one call runs search as the required leg and research-brief in parallel as an optional leg, so an LLM/scrape hiccup degrades the response instead of failing the call. Public-source data snapshot only — not an ESG rating, score, or compliance opinion. Use it as an ESG posture scan or public sustainability-disclosure digest for agents doing corporate research. | | filing-financials-brief | (0.07 USDC/call) Get an SEC filing brief in one call instead of stitching three EDGAR lookups together: the filing itself, its XBRL financials, and a plain-language filing summary. Buyers searching for a sec filing brief, xbrl financials, company financials extract, or 10-K summary get filing metadata and text, an XBRL revenue/net income/EPS/assets series, and an executive-style summary in one response. Composite: one call runs sec-filing + xbrl-financials in parallel, then summarize-text (optional, LLM-backed) once filing text resolves, merged into filing, financials, and summary sections. Public-data extraction from SEC EDGAR only: no investment advice, no ratings, no trade recommendations, just what the company disclosed. Accepts a ticker or CIK plus an optional form_type (10-K default, 10-Q, 8-K, etc). Use it as a filing summary API, an EDGAR research shortcut, or a due-diligence financials puller. | | flight-brief | (0.03 USDC/call) Reports live flight status with FAA airport delays and airport info in one call. Composite: one call runs flight-status + airport-delays + iata-airport-info. Live aircraft state for a flight callsign (OpenSky Network) plus FAA ground stops, GDPs, and delay programs. Pass optional origin/destination IATA-3 codes to get per-airport delay events and airport metadata (name, coordinates, links); omit them and the national FAA delay picture is attached instead. Live data from OpenSky Network and FAA NAS Status. Use it as a flight tracker bundle or callsign lookup with delay context. | | game-preview-brief | (0.05 USDC/call) Neutral pregame brief for a matchup: recent results and upcoming schedule for the named teams from a major-league schedule feed, plus recent news/X coverage, in one call. Composite: one call runs sports-schedule twice (scope=recent and scope=upcoming, filtered to the named teams) as the required legs plus search as an optional third leg. Neutral stats and schedule framing only — no betting lines, picks, spreads, or wagering guidance anywhere in the output. Use it as a pregame research brief or matchup schedule-and-news bundle for agents tracking upcoming games. | | grant-finder | (0.05 USDC/call) Structured grant and funding-opportunity finder for a field and region: live federal opportunities from grants.gov's index plus a cited, synthesized web answer covering broader, international, or private funding grants.gov doesn't index, in one call. Composite: one call runs grant-lookup as the required leg and answer-web in parallel as an optional leg, so an LLM/scrape hiccup on the broader summary degrades the response instead of failing the call. Use it as a grant/funding-opportunity scan for researchers, nonprofits, and small businesses scouting money on the table. | | grounded-answer-check | (0.06 USDC/call) Answer a question and grade how well the answer actually holds up against independent sources, in one call. Composite: one call runs answer-web + web-search + wikipedia-search + retrieval-rerank + citation-verify in parallel, then maps the answer's sentences onto reranked, re-verified evidence passages. Pass a question to get a generated, cited answer, or pass your own answer as context to grade it instead. Returns a support score, a supported/unsupported verdict, and the exact spans that had no backing evidence. Use it as a grounded answer API, hallucination check, RAG grounding evaluator, or source-backed QA layer before an agent trusts and repeats an answer. | | image-asset-pipeline | (0.08 USDC/call) Turns a text prompt into a production-ready image asset in one call: generate, sharpen, cut out the background. Composite: one call runs image-generate, then image-upscale, then remove-bg in sequence, merging into { image_url, stages }, with no separate paid calls or URL hand-offs between endpoints. Covers product image generation, ecommerce catalog creative, and one-call image workflow needs where a buyer wants a finished transparent-PNG asset, not a raw generation. Each downstream stage is optional and degrades gracefully: if upscale or background removal fails (or is disabled via input flags), the prior stage's image carries forward instead of failing the whole call. | | image-caption-localize | (0.04 USDC/call) Captions an image and translates the caption into any of 100+ languages in one call. Composite: one call runs describe-image + translate. A vision LLM writes a single-sentence caption, then it is translated into the target language. Returns the source caption, translated caption, and per-component telemetry. Use it for translated image captions, multilingual alt text, or vision caption plus translation. | | image-generate-localize | (0.04 USDC/call) Generates an image from a text prompt and returns a localized caption to go with it in one call. Composite: one call runs image-generate + translate in parallel. Built for agents assembling a social post image, a multilingual creative asset, or a localized ad variant where a caller would otherwise fire off a text-to-image API and a translate API as two separate paid calls. Send a prompt and a target_language (optionally a separate caption if the on-image text should differ from the generation prompt); get back the generated image URL plus the caption in both the original text and the localized translation. Image generation is required; translation degrades the response instead of failing the call if its upstream is slow or down. Use it as an image generation API with built-in localization, a social post image generator, or a multilingual creative asset tool. | | image-intel-pack | (0.04 USDC/call) Analyzes an image in one call: description, brand logo detection, and content moderation. Composite: one call runs describe-image + logo-detect + moderate-content. Vision LLM description and logo detection run in parallel, then the description text is moderated for safety (harassment, hate, violence, sexual content, spam, phishing). Returns description, detected brands, moderation verdict, and per-component telemetry. Use it as an image analysis bundle or vision intelligence pack. | | inbound-message-firewall | (0.06 USDC/call) One call an agent makes before trusting an inbound message that also carries a claimed source link: an email with a link, a DM citing a URL, a tool result referencing a page. Composite: one call runs prompt-injection-detect (required, a failure fails the call), moderate-content and ai-content-detector (both degradable, the latter skips silently under its own length floor), and content-authenticity-report on source_url when given (degradable). agent-input-guard screens text only; this adds the source-URL check, so an agent can decide whether to act on 'per this report, do X' in one call. Returns one act/verify/block decision, a risk score, and per-leg detail. Use it as an inbound message firewall, agent trust gate, message-plus-source verification API, or a pre-action guardrail for autonomous agents. | | json-dev-pack | (0.03 USDC/call) Converts a JSON document to TypeScript types and YAML, with optional schema validation, in one call. Composite: one call runs type-inference-from-json + json-yaml in parallel, plus json-schema-validate when you supply a schema. Returns a generated TypeScript interface (with date-time/uuid/uri/email format detection), the YAML rendering of the same document, and, when a JSON Schema (draft-07 or 2020-12) is provided, a validation report with per-error instance_path + keyword + message. Input capped at 200k chars. Includes composed_of + per-component telemetry; partial failure degrades instead of failing. Use it as an API response toolkit, quicktype-style type generation with format conversion, or structured-output verifier bundle. | | lead-qualify | (0.06 USDC/call) Qualifies a lead in one call: email validity, disposable check, person enrichment, and company enrichment. Composite: one call runs lead-enrich + email-deliverability-check + disposable-email-check + company-enrich in parallel and merges them into a qualification object {email_valid, disposable, person, company} plus a deterministic fit_signals array (e.g. corporate_email_domain, domain_accepts_mail, disposable_email_domain). Data and derived signals only; no lead scoring, ranking, or buying advice. Company context is derived from the email domain when no company name is supplied. Use it as a lead qualification data API, lead verification, signup vetting, or CRM intake check for sales and growth agents. | | literature-landscape | (0.06 USDC/call) Prior art search, literature review, and grant discovery in one call: search arXiv papers, PubMed biomedical literature, USPTO patents, and grants.gov funding opportunities for a single scholarly or IP query, then dedupe the same work across corpora and verify the top hit. Composite: one call runs arxiv-search, pubmed-search, patent-search, grant-lookup, and citation-verify in parallel and merges results into one cross-corpus landscape. Send a query and an optional include list of corpora (papers, biomedical, patents, grants) plus a per-corpus limit; get back results grouped by corpus, a deduped merged list, counts, and a citation check on the top hit. Use it as a prior-art search API, an academic research aggregator, or a grant discovery tool for scholarly sweeps. | | localize-pack | (0.05 USDC/call) Localizes text in one call: detects the source language, translates, and optionally rewrites the translation in a chosen tone. Give it text and a target language and get back the detected source language (ISO code + confidence), a fluent translation, and, when a tone is supplied, the translated text rewritten in that tone (formal, casual, friendly, persuasive, and 8 more). One paid call replaces the detect, translate, restyle chain: ideal for support reply localization, multilingual publishing, agent-to-human messaging. Composite: one call runs detect-language + translate in parallel, then rewrite-tone over the translation when a tone is provided. Use it as a localization pipeline API or multilingual content adapter. | | macro-snapshot | (0.05 USDC/call) Daily macro monitor: major-pair FX rates against USD, a recency-scoped web search for macro headlines (optionally scoped to a region), and a structured research brief on the macro backdrop, in one call. Composite: one call runs exchange-rates as the required leg, then search and research-brief in parallel as optional legs, so an LLM or scrape hiccup degrades the response instead of failing the call. Informational monitoring only — no investment recommendations. Use it as a daily macro/FX monitor for agents tracking a region's economic backdrop. | | market-mindshare-monitor | (0.05 USDC/call) Re-runnable mindshare monitor for a topic: cross-platform social-mindshare (X, news, Hacker News), any live Polymarket prediction markets referencing the topic, and a general web search pass for recent coverage, in one call. Composite: one call runs social-mindshare as the required leg, then polymarket-markets and search in parallel as optional legs, so a quiet prediction-market or search result degrades the response instead of failing the call. Data-only Polymarket leg: no trading, order execution, or betting guidance. Use it as a recurring mindshare and prediction-market monitor for brands, tickers, or topics. | | market-rates-pack | (0.04 USDC/call) Merges FX rates, perp funding, and Hyperliquid mark price into one call. Composite: one call runs exchange-rates + funding-rates + hyperliquid-market in parallel, merging ECB fiat FX rates for a chosen base currency, Binance USD-M perpetual funding (current rate, mark/index price, next funding time, 24h history), and Hyperliquid perp market data (mark price, max leverage, size decimals) for a chosen crypto symbol. Data-only market data merge; no advice. Use it as a cross-market rates bundle: ECB exchange rates + Binance funding + perps data. | | materials-price-check | (0.05 USDC/call) Supply/materials price check: shopping-intent retail listings with detected prices, a web search for bulk/wholesale pricing context, and (when a non-USD currency is requested) an fx reference rate, in one call. Composite: one call runs product-search as the required leg, then search and (when needed) exchange-rates in parallel as optional legs, so a miss on either degrades the response instead of failing the call. Use it as a materials or supply price check for agents doing procurement or cost research. | | mcp-server-safety-report | (0.06 USDC/call) MCP server audit / mcp safety report / tool risk review / agent tool security in one call: point it at a live MCP server URL or hand it a tool manifest and get a risk read before an agent starts calling that server's tools. Composite: one call runs mcp-tools-list + mcp-tool-risk-score + prompt-injection-surface + secrets-exposure-check in parallel, merged into {risk, injection_surface, secrets_exposure}. mcp-tools-list discovers the tool catalog from server_url via JSON-RPC tools/list; mcp-tool-risk-score scores each tool 0-100 for permission scope and destructive-action risk. Pass a GitHub repo alongside the server to also scan the server's own code for prompt injection and secrets exposure. Use it as an MCP server audit API, an agent tool security checker, or a pre-connect safety gate. | | meeting-brief | (0.08 USDC/call) Turns a meeting or call recording into speaker-labeled notes and a summary in one call. Composite: one call runs speaker-diarize + summarize-text. Diarizes the recording (Whisper v3 + speaker labels), then summarizes the speaker-labeled transcript. Returns utterances by speaker, speaker count, labeled transcript, summary, and per-component telemetry. Use it for meeting summaries from recordings, who-said-what digests, or diarized meeting notes. | | mention-intel | (0.06 USDC/call) Mention intelligence API: send a handle, brand, ticker, or phrase and get back recent X/Twitter mentions with each author's follower count, verification status, and bot-likelihood score, plus an overall sentiment read across the mention text. Built for the workflow agents already run by hand: pull mentions, then separately enrich the authors that matter, then eyeball the tone. Composite: one call runs mentions + twitter-user-lookup + x-account-authenticity + sentiment-analysis. Mentions search is required; per-author enrichment (capped at the 5 most distinct authors) and the batch sentiment pass degrade the response instead of failing the call on a rate limit or LLM hiccup. Use it as a mention monitoring API, brand/handle mention tracker, or X mentions-plus-author-enrichment tool for agents doing social listening or reputation triage. | | model-selection-brief | (0.05 USDC/call) Model selection / route and cost brief / which model / prompt cost in one call: describe a task and get a recommended model tier, candidate models, and, when you pass a sample prompt, its token count, a projected cost across those candidates, and a compressed version of the prompt. Composite: one call runs model-route-recommend + llm-cost-estimate + token-count + prompt-compress, merged into {route, token_count, cost_estimate, compressed_prompt}. model-route-recommend picks a tier and named models from the task and optional latency/budget/context constraints; token-count and llm-cost-estimate price that exact prompt against the recommended models; prompt-compress shortens the same prompt. Without a sample prompt, only routing runs. Use it as a model selection API, an LLM routing advisor, or a which-model-should-I-call tool. | | multilingual-doc-brief | (0.06 USDC/call) Turn a foreign-language document into a plain-English brief in one call: send a PDF and get back extracted text, the detected source language, an English (or target-language) translation, a summary, and a PII risk flag. Built for agents that need doc translation, a multilingual document brief, or to translate and summarize a foreign doc without chaining five separate calls themselves. Composite: one call runs pdf-to-text first, then detect-language + translate-text + summarize-text + detect-pii in parallel on the extracted text, and merges the results into {text, detected_language, translation, summary, pii}. Send a PDF URL (url or file) and an optional target_lang (default 'en'). Enrichment legs are LLM-backed and degrade gracefully if one fails; extraction failure fails the call. | | naming-pack | (0.06 USDC/call) Runs a full startup naming workflow in one call: generates brandable candidates, then checks domains, brand score, and social handles. Composite: one call runs business-name-generator first, then fans out domain-availability, brand-name-score, and social-handle-check over the top candidates (bounded to 9 component calls total). Returns brainstormed candidates ranked by available TLDs + brandability, then an enriched shortlist: brand-quality score with risks and recommendations, X/Twitter handle availability, and RDAP .com registration status with registrar and expiry. Enrichment checks are optional; failures degrade the response instead of failing it. Includes composed_of + per-component telemetry. Use it as a business name generator, company naming bundle, or brandable domain finder. | | news-content | (0.05 USDC/call) News-with-full-text API: send a topic and get back recent news articles with each article's cleaned page text already attached, instead of a headline list you have to fetch yourself. Composite: one call runs news-search, then scrapes and cleans the top articles' page text in parallel, capped at 4,000 characters per article. news-search is required; each article fetch is an independent optional leg, so a slow or blocked page returns content: null with a note on that one article instead of failing the whole call. Use it as a news API with article content, a news-and-extract tool, or a recent-news fetch for agents that need article text, not just headlines. | | news-digest | (0.05 USDC/call) Summarizes the news on any topic in one call: ranked results, a bullet digest of the coverage, and the named entities (people, companies, places) driving the story. Give it a topic and a recency window (day/week/month). One paid call replaces a search-then-summarize agent loop: ideal for daily briefings, brand monitoring, market awareness, research agents. Composite: one call runs web-search first, then summarize-text + extract-entities in parallel over the result snippets. Use it as a topic news summary, current events briefing, or agent news monitor. | | npm-trust-pack | (0.04 USDC/call) Vets an npm package before install in one call: supply-chain risk score plus registry stats. Composite: one call runs package-risk-npm + npm-package-stats in parallel for any npm package (scoped names supported). Returns the 0-10 risk score with risk_level bucket, contributing factors, typosquat candidates, and plain-English summary, plus the raw registry metrics: latest version, version count, age, days since last release, maintainer list, and last day/week/month/year downloads. Includes composed_of + per-component telemetry; partial failure degrades instead of failing. Use it as an npm install pre-flight, typosquat check + downloads + maintainers bundle, or Socket.dev-adjacent vetting bundle. | | onchain-identity-pack | (0.03 USDC/call) Resolves an EVM address or name across ENS, Basenames, and known-wallet labels in one call. Composite: one call runs ens-resolve + basename-resolve + wallet-label in parallel. Pass a 0x address for reverse ENS, reverse Basename, and known-address labelling, or pass a name ('vitalik.eth', 'jesse.base.eth', bare 'alice') for forward resolution on both name services with text records. All components are optional individually; at least one must resolve or the call returns 424. Data-only aggregation; no advice. Use it as a name-to-address and address-to-name resolver bundle, vitalik.eth lookup, .base.eth lookup, or EVM address identity data. | | outreach-pack | (0.05 USDC/call) Drafts a cold email and verifies the recipient address in one call. Composite: one call runs email-draft + email-deliverability-check in parallel. Returns a ready-to-send email (subject, body, salutation, sign-off; tone controllable) plus a deliverability report on the recipient address: syntax, MX records, disposable-domain and role-account detection, SPF/DMARC/DKIM posture, and a 0-100 deliverability score. Privacy-safe DNS only, no mailbox probing. Includes composed_of + per-component telemetry; partial failure degrades instead of failing. Use it as a sales outreach pre-flight or inbox-readiness bundle. | | paper-brief | (0.04 USDC/call) Finds and digests an arXiv paper in one call: structured summary plus a ready-to-paste BibTeX citation. Composite: one call runs arxiv-search + arxiv-summarize + arxiv-bibtex. Pass a topic query (resolves to the top relevance hit) or an arXiv ID (search stage skipped); the structured LLM summary (TLDR, key findings, methods, limitations) and the BibTeX entry are generated in parallel. Public arXiv API. Use it as an arXiv paper digest, research paper summary, or search-summarize-cite bundle. | | patent-landscape-brief | (0.08 USDC/call) Approximate patent/prior-art landscape for a topic, built from published technical prior art rather than a patent-office filing search: arXiv full-text search results, a web search scoped to filings and patent news, and a research brief tying it together, in one call. Composite: one call runs arxiv-search + search in parallel (both required together — the call only fails if both fail) plus research-brief as an optional third leg. Not a patent-office prior-art search or freedom-to-operate opinion. Use it as a prior-art landscape scan or patent-adjacent research brief for agents doing IP due diligence groundwork. | | person-dossier | (0.05 USDC/call) Researches a person from their email address in one call: enrichment, deliverability, disposable check, and social handle. Composite: one call runs person-enrich + email-deliverability-check + disposable-email-check + social-handle-check in parallel and merges them into {person, deliverability, disposable, social_handle}. The email local-part is probed as a candidate X handle (optional, degrades gracefully). Public-data enrichment only; no private people databases, no social-network scraping. Use it as a person dossier API, contact research, email intelligence, or lead profile lookup for CRM and sales agents. | | pr-pack | (0.05 USDC/call) Turns a git diff into a commit message and PR description in one call. Composite: one call runs commit-message-from-diff + pr-description-from-diff in parallel over the same unified diff. Returns a Conventional Commits message (type, scope, subject, full_message) plus a review-ready PR description (title, summary bullets, Markdown test-plan checklist, breaking_change flag, paste-ready body). Diff capped at 30,000 chars. Response includes composed_of + per-component telemetry; partial upstream failure degrades instead of failing. Use it as a pull request bundle, conventional commit + PR body generator, or ship-a-branch helper. | | prediction-market-snapshot | (0.04 USDC/call) Searches Polymarket and Kalshi prediction markets with one query, returning both venues' open markets side by side. Composite: one call runs polymarket-markets + kalshi-markets in parallel with the same free-text query and returns merged {polymarket: [], kalshi: []} lists: questions, tickers, implied outcome prices (0-1), volume, liquidity, and end dates from both venues. Data-only listing of public market data; no probability advice, no order entry, no trading. Use it as a cross-venue market search or event odds listing bundle. | | product-listing-pack | (0.06 USDC/call) Preps an ecommerce product listing in one call: marketing copy, SEO keywords, and a product-name quality score. Composite: one call runs product-description + keyword-suggest + product-name-score in parallel. Returns marketing-ready copy (description, selling points, key features, listing titles built from your supplied facts), a de-duplicated ranked SEO keyword list from Wikipedia + DuckDuckGo autocomplete, and a name-quality score with grade, risks, and recommendations. name-score is optional: its failure degrades the response instead of failing it. Includes composed_of + per-component telemetry. Use it as a marketplace listing generator or Shopify / Amazon / Etsy listing prep. | | product-purchase-brief | (0.05 USDC/call) Product purchase brief API for a buy decision: send a raw product title (optionally with return-policy and warranty text) and get back normalized product attributes, extracted return and warranty terms, and a short description in one call. Built for agents comparing listings so they skip running a SKU normalizer, a return-policy extractor, and a warranty parser as three separate calls. Composite: one call runs sku-normalize + return-policy-extract + warranty-parse + product-describe in parallel, merged into {normalized, return_policy, warranty, description}. sku-normalize is required; product-describe is optional and can time out; return-policy-extract and warranty-parse only run when policy_text/warranty_text are supplied. Use it as a product purchase brief API, buy-decision helper, or return-and-warranty summarizer. | | property-diligence-pack | (0.05 USDC/call) Property diligence API for a street address: geocode, tax assessment, permit history, seismic hazard, and air quality in one call for real estate risk screening. Composite: one call runs address-geocode first to resolve the address, then property-tax-assessment, building-permits, earthquakes-recent, and air-quality in parallel, merged into one address risk profile. Tax assessment and permit history only cover a handful of US open-data cities and degrade gracefully elsewhere; seismic and air-quality legs run for any resolved coordinate. Use it as a property tax lookup, permit history check, seismic hazard screen, or address risk profile for real-estate diligence. | | pubmed-brief | (0.04 USDC/call) Searches PubMed and summarizes the matching biomedical literature in one call. Composite: one call runs pubmed-search first, then summarize-text over the retrieved titles + abstracts. Returns the matching papers (PMID, title, authors, journal, date, abstract, DOI) plus a bullet-point roll-up of what the abstracts say. Literature summaries only; this is NOT medical advice, and output describes published research and always carries a consult-a-clinician disclaimer. Summary stage is optional: its failure degrades the response instead of failing it. Federal public-domain data via NCBI E-utilities. Includes composed_of + per-component telemetry. Use it as an NIH paper digest, research literature synthesis, or MeSH query with abstract roll-up. | | pypi-trust-pack | (0.04 USDC/call) Vets a Python package before pip install in one call: supply-chain risk score plus package stats. Composite: one call runs pypi-package-risk + pypi-package-stats in parallel for any Python package. Returns the 0-10 risk score with risk_level bucket, contributing factors, typosquat candidates, and plain-English summary, plus the raw PyPI metrics: latest version, version count, age, days since last release, author/maintainer, classifiers, and pypistats day/week/month downloads. Includes composed_of + per-component telemetry; partial failure degrades instead of failing. Use it as a pip install pre-flight, typosquat check + downloads + release history bundle, or requirements.txt + pyproject.toml vetting bundle. | | receipt-to-ledger | (0.05 USDC/call) Turn a receipt photo into a ledger entry in one call: line items, merchant and date entities, and a currency-normalized total, ready for a bookkeeping or expense system. Composite: one call runs receipt-parser first, then extract-entities and exchange-rates in parallel off the OCR output, merging into { ocr, line_items, entities, totals }. Covers receipt to ledger conversion, receipt parsing, expense extraction, and receipt ocr structured output for accounting and expense-tracking agents. Send image_url or file (receipt photo URL) and an optional target_currency (default USD); totals.converted carries the same-day FX-normalized amount. Use it as a receipt parsing API, expense automation tool, or receipt-to-ledger pipeline for finance agents. | | regex-pack | (0.03 USDC/call) Generates a regex from a natural-language prompt and live-tests it in one call. Composite: one call runs regex-from-prompt first, then feeds the generated pattern into regex-test over your test_strings (optional stage, runs only when test_strings are provided). Returns the JavaScript pattern + flags + explanation + examples, plus per-input matched/not-matched verdicts with capture groups so you know the pattern actually matches what you asked for. Prompt capped at 2,000 chars; up to 100 test strings of 5,000 chars each. Includes composed_of + per-component telemetry. Use it as an NL pattern builder with verification, generate-and-verify regex workflow, or regex playground bundle. | | repo-health-brief | (0.05 USDC/call) Assesses whether a public GitHub repo is safe to depend on, in one call: health score, README, and star history. Composite: one call runs github-repo-health + github-readme + github-stars-history in parallel. Returns the 0-100 maintainability score with grade (abandoned/stale/okay/healthy/thriving) and LLM verdict, the raw README markdown (truncated at 20k chars), and a sampled cumulative-stars-vs-date growth series. stars-history is optional: if it fails the response degrades instead of failing. Includes composed_of + per-component telemetry. Use it as a GitHub repo brief, open-source dependency vetting bundle, or should-I-depend-on-this-repo check. | | research-radar | (0.05 USDC/call) Re-runnable topic monitor: the latest arXiv papers sorted by most-recently-updated, recent (week-scoped) web coverage, and a structured research brief, in one call. Composite: one call runs arxiv-search + search in parallel (both required together — the call only fails if both fail) plus research-brief as an optional third leg. Distinct from patent-landscape-brief, which sorts arXiv by relevance and frames the same two search legs around prior art rather than what's newly published. Use it as a recurring topic-monitoring feed for agents tracking a fast-moving research area. | | research-topic-scan | (0.06 USDC/call) Research scan API for a multi source search sweep: send one query and get back web results, recent news, an encyclopedia summary, academic papers, developer/Hacker News chatter, and X/Twitter chatter in a single call. Built for agent-led research, topic research, and due-diligence sweeps where a caller would otherwise fire off six separate lookups and merge them by hand. Composite: one call runs web-search + news-search + wikipedia-search + arxiv-search + hn-search + x-search in parallel, dedupes hits by URL, and tags each result with its source. Only web-search is required; news/wikipedia/arxiv/hn/x-search degrade the response instead of failing the call if their upstream is slow or down (x-search also needs a configured X bearer token). Use it as a research scan API, a topic research tool, or a web-news-papers-tweets aggregator for agents scoping a subject before they write or decide. | | resume-review-pack | (0.06 USDC/call) Reviews a resume in one call: ATS fit score, PII exposure scan, and a summary polish. Composite: one call runs resume-scorer + detect-pii + rewrite-tone in parallel. Returns the 0-100 ATS score with calibrated subscores (keyword match, experience, skills, formatting, impact) and ranked improvement suggestions, a PII exposure report (emails, phones, addresses, SSNs) with risk level, and a confident-tone rewrite of the resume's summary/objective section. The rewrite stage is optional and bounded; its failure degrades the response instead of failing it. Includes composed_of + per-component telemetry. Use it as a job application pre-flight, resume vs job description fi