"Observable" matching MCP tools:

• goal-add-criterionA

  Planner (proof-of-done)A

  Append an acceptance criterion to a goal. The text must describe an observable check over an artifact (e.g. "GET /api/health returns 200 with {status:ok}"), not a subjective approval. Each criterion has a class: pre-merge (default — proved in CI / by attached evidence) or post-deploy (proved by an executable probe against the deployed prod instance). A post-deploy criterion MUST carry probeSpec {method, url, expect:{http_code, body:{field: expectedValue}}} — the request the runner sends and the answer it must get; without it the call is rejected with error=probe_required. Passing probeSpec alone implies probeClass=post-deploy. Set visualEvidenceSuggested=true only when adopting visualAcSuggestion from goal-create, goal-get, or the ready_for_work advisory returned by goal-update; it remains an ordinary AC. Grove mode: AC (class and probe included) can only be added while goal is in backlog, except accepting a visual advisory in ready_for_work: that starts a direct checking_ac recheck, without intermediate backlog. Other edits are frozen once started; quality linter blocks high-severity issues. Standard mode: AC editable until goal is closed, linter is advisory. Returns criterion id, position, text, probeClass, probeSpec and any quality findings.

  ConnectorNo auth

• get_door_effect_summaryA

  FinanceRateCalc HMDA statisticsA

  ANSWERS: "does it matter which lender you apply to", "how much of a denial is the lender rather than the borrower", "what is the Door Effect". Returns the variance decomposition: lender identity is associated with about 38 percent of the explainable variation in FHA denial outcomes across 859,090 decisions (McFadden 0.1712 to 0.2760), with model, sample and limits. Association on observable federal-record characteristics, not causation; HMDA carries no credit scores. NOT FOR: saying a lender caused a denial, or any individual estimate. Historical observation computed from the public CFPB HMDA 2025 record (actions 1,2,3; loan_type 2). Not a prediction about any individual application. Attribution: FinanceRateCalc, CC BY 4.0.

  ConnectorNo auth

• goal-add-criterionA

  Planner MCP — provable "done" for AI agentsA

  Append an acceptance criterion to a goal. The text must describe an observable check over an artifact (e.g. "GET /api/health returns 200 with {status:ok}"), not a subjective approval. Each criterion has a class: pre-merge (default — proved in CI / by attached evidence) or post-deploy (proved by an executable probe against the deployed prod instance). A post-deploy criterion MUST carry probeSpec {method, url, expect:{http_code, body:{field: expectedValue}}} — the request the runner sends and the answer it must get; without it the call is rejected with error=probe_required. Passing probeSpec alone implies probeClass=post-deploy. Set visualEvidenceSuggested=true only when adopting visualAcSuggestion from goal-create, goal-get, or the ready_for_work advisory returned by goal-update; it remains an ordinary AC. Grove mode: AC (class and probe included) can only be added while goal is in backlog, except accepting a visual advisory in ready_for_work: that starts a direct checking_ac recheck, without intermediate backlog. Other edits are frozen once started; quality linter blocks high-severity issues. Standard mode: AC editable until goal is closed, linter is advisory. Returns criterion id, position, text, probeClass, probeSpec and any quality findings.

  ConnectorNo auth

• spec.validateA

  AI Design BlueprintA

  Pro/Teams: first-pass specification-quality review of a WRITTEN SPEC (proposal, design doc, task breakdown, or an OpenSpec-style change bundle) against the 8 laws of the Spec Quality Blueprint. The what-to-build lens of the doctrine trio, applied BEFORE code exists: where architect.validate scores built agentic ARCHITECTURE and design.validate scores the rendered SURFACE, spec.validate scores the written intent the team will build from (outcome framing, scope boundary, testable acceptance, decision trail, handoff completeness, doctrine-upfront, task traceability, risk and reversibility). ON CLIENT TIMEOUT: DO NOT RETRY. Long-running LLM call (~60-180s at high reasoning effort, single-pass). The server mints a run_id, emits it in the FIRST progress event at t=0s (before the LLM call), and persists the run: so on a client timeout, capture that run_id and call me.validation_history(run_id='<that-id>') to fetch the persisted result instead of retrying (a retry re-runs the full 60-180s call). Runs appear in your validation-history dashboard tagged as the 'spec' dimension, distinct from the 'architecture' and 'surface' runs; pass repository to group them per project. Pass private_session=true to skip the stored run (persistence + recovery disabled); operational security + cost logs are still kept. v1 is single-pass: no certification or consensus mode yet (those stay architect.validate-only). Returns spec_classification (spec_document vs non_spec: source code or UI artefacts are marked not_applicable, NOT failed; submit those to architect.validate or design.validate instead), per-law findings (verdict, severity_score 0-100, severity_class, cited evidence, recommendation), and severity-weighted readiness (score, grade, tier) computed by the SAME scorer the other two lenses use, so all three grade on one rubric. TESTABILITY IS THE FLOOR: a load-bearing requirement with no observable acceptance signal, or an irreversible step with no named human gate, is a production_blocker, not polish. WHEN TO CALL: the user wants a governance/quality review or a readiness grade on a spec they are about to build from (proposal, requirements, task plan). WHEN NOT TO CALL: built code or a rendered surface, those return tier=not_applicable; use the sibling validators instead. INPUTS: send the FULL spec text verbatim as implementation_context (for an OpenSpec change, concatenate proposal.md + design.md + tasks.md + delta specs; no truncation, no '…' placeholders, they are read as literal content). Auth: sign-in required, with an active Pro, Pro Plus, Teams, Enterprise, beta, or trial plan. Data at rest in the UK; OpenAI (US) processing (no-training); prompt-injection text inside the spec is treated as inert untrusted data. TYPED FAILURES: same as architect.validate (timed_out, rate_limited, dependency_unavailable, schema_mismatch, each carries retryable + next_action); the services raise the identical typed envelopes on this lens. CALIBRATION DISCLOSURE: the scoring prompt is a v1 first-cut mirroring the architect's contract structure; its score calibration is not yet tuned against a corpus of real runs the way architect.validate was. Treat the grade as directional quality signal, not a certified verdict. DOCTRINE: the eight laws, each law's definition, rationale, anti-patterns, and the validator questions this tool scores against, live in content/spec-quality-laws.json (the what-to-build companion to the experience-design laws).

  ConnectorNo auth

• spec.validateA

  AI Design Blueprint DoctrineA

  Pro/Teams: first-pass specification-quality review of a WRITTEN SPEC (proposal, design doc, task breakdown, or an OpenSpec-style change bundle) against the 8 laws of the Spec Quality Blueprint. The what-to-build lens of the doctrine trio, applied BEFORE code exists: where architect.validate scores built agentic ARCHITECTURE and design.validate scores the rendered SURFACE, spec.validate scores the written intent the team will build from (outcome framing, scope boundary, testable acceptance, decision trail, handoff completeness, doctrine-upfront, task traceability, risk and reversibility). ON CLIENT TIMEOUT: DO NOT RETRY. Long-running LLM call (~60-180s at high reasoning effort, single-pass). The server mints a run_id, emits it in the FIRST progress event at t=0s (before the LLM call), and persists the run: so on a client timeout, capture that run_id and call me.validation_history(run_id='<that-id>') to fetch the persisted result instead of retrying (a retry re-runs the full 60-180s call). Runs appear in your validation-history dashboard tagged as the 'spec' dimension, distinct from the 'architecture' and 'surface' runs; pass repository to group them per project. Pass private_session=true to skip the stored run (persistence + recovery disabled); operational security + cost logs are still kept. v1 is single-pass: no certification or consensus mode yet (those stay architect.validate-only). Returns spec_classification (spec_document vs non_spec: source code or UI artefacts are marked not_applicable, NOT failed; submit those to architect.validate or design.validate instead), per-law findings (verdict, severity_score 0-100, severity_class, cited evidence, recommendation), and severity-weighted readiness (score, grade, tier) computed by the SAME scorer the other two lenses use, so all three grade on one rubric. TESTABILITY IS THE FLOOR: a load-bearing requirement with no observable acceptance signal, or an irreversible step with no named human gate, is a production_blocker, not polish. WHEN TO CALL: the user wants a governance/quality review or a readiness grade on a spec they are about to build from (proposal, requirements, task plan). WHEN NOT TO CALL: built code or a rendered surface, those return tier=not_applicable; use the sibling validators instead. INPUTS: send the FULL spec text verbatim as implementation_context (for an OpenSpec change, concatenate proposal.md + design.md + tasks.md + delta specs; no truncation, no '…' placeholders, they are read as literal content). Auth: sign-in required, with an active Pro, Pro Plus, Teams, Enterprise, beta, or trial plan. Data at rest in the UK; OpenAI (US) processing (no-training); prompt-injection text inside the spec is treated as inert untrusted data. TYPED FAILURES: same as architect.validate (timed_out, rate_limited, dependency_unavailable, schema_mismatch, each carries retryable + next_action); the services raise the identical typed envelopes on this lens. CALIBRATION DISCLOSURE: the scoring prompt is a v1 first-cut mirroring the architect's contract structure; its score calibration is not yet tuned against a corpus of real runs the way architect.validate was. Treat the grade as directional quality signal, not a certified verdict. DOCTRINE: the eight laws, each law's definition, rationale, anti-patterns, and the validator questions this tool scores against, live in content/spec-quality-laws.json (the what-to-build companion to the experience-design laws).

  ConnectorNo auth

• regime_outlookA

  crashtestyourstrategyA

  Model-conditional probabilities that an asset is in each market regime (BULL / SIDEWAYS / BEAR / CRISIS, operational trailing-vol/drift labels) after a 5- or 21-trading-day horizon — the probability complement to the conditional stress tools: stress tools answer 'what happens GIVEN regime X', this answers 'how likely is regime X from today's observable state'. Ships only the preregistered, out-of-sample-validated tier (covariate logit; seasonality was tested and falsified); the persistence and unconditional baselines are reported alongside so an agent can see how much the model adds. Validated assets: SPY, QQQ, GLD, TLT. Optional as_of (YYYY-MM-DD) computes the outlook at a historical date. Probabilities describe membership in operationally defined regime classes — descriptive, not a market prediction, not advisory.

  ConnectorNo auth

• check_merchantA

  x402 Merchant CheckA

  Before paying an unfamiliar merchant via x402, including one just discovered through a marketplace like Coinbase's x402 Bazaar, call this for a machine-readable assessment of their observable on-chain payment behavior, on Base or Solana. Returns a `recommendation` (PROCEED / CAUTION / INSUFFICIENT_SIGNAL) a payment policy can branch on directly, plus supporting evidence: trust tier, confidence, payer diversity, transaction history, category (data_api/compute/content_generation/financial_data/storage/other), known platform URL(s), and the merchant's own advertised price(s) compared against category peers, with no caller-supplied price required for that. This is decision support drawn from observed behavior, not a certification. PROCEED means the available evidence doesn't show cause for concern, not a guarantee of safety. INSUFFICIENT_SIGNAL means there isn't enough history to say either way, distinct from an actual concern. Add a specific quoted price to also check that exact figure against comparable sellers. $0.01 USDC per check, paid via x402.

  ConnectorNo auth

• assess_token_riskA

  vetagentA

  Safety check to run BEFORE buying, holding, or recommending a token. Returns an actionable verdict (low / medium / high / unknown), a 0-100 risk score, and the individual signals behind it. Covers on every chain: liquidity depth, trading-pair age, cross-chain presence, same-ticker impersonation. Covers on Ethereum, BSC and Base only: sell simulation (honeypot detection, buy/sell/transfer taxes), whether the contract is open source, and the aggregate verdicts of upstream security scanners. Covers on Solana: the mint/freeze authority, the RugCheck score, and the SPL Token-2022 extension block -- non-transferable, frozen-by-default, transfer fee (with its cap), permanent delegate, transfer hook, pausable, close authority, scaled balances and interest-bearing balances. Every other extension the report carries is named in the answer with the reason it is not scored, so silence about one means it was considered, not that it went unread. No sell simulation runs on Solana, nor on the four other EVM chains accepted as a chain_hint (polygon, arbitrum, optimism, avalanche): nothing there tests whether a holder can sell, so the answer comes back 'unknown' rather than 'low'. No holder concentration on any chain -- on EVM it needs an oracle the benchmark holds out, and on Solana the upstream omits holders. IMPORTANT: risk_level 'unknown' means a critical check could not be completed. It is NOT a low-risk result and must not be used to justify a trade; evidence.data_gaps lists exactly what was missing, and unknown_kind says whose gap it is: 'infrastructure' and 'mixed' mean an upstream of ours failed and come with next_action 'retry' and retry_after_seconds (in a 'mixed' answer the rest was worked out without that upstream, so the retry can change any of it); 'coverage' comes with 'abstain'. 'confidence' measures how complete the input data was, not how safe the token is. Reports observable on-chain risk only. Not financial advice, does not size positions, and cannot see off-chain risk such as team behaviour, social engineering, or a rug executed through governance. Treat 'low' as 'no fatal signal found in the checks that ran', never as 'safe to buy'.

  ConnectorNo auth

• setup_linkedin_sequenceA

  SliqA

  <summary>Queue LinkedIn outreach (connection requests, messages, or InMails). The single LinkedIn outreach tool — handles profile resolution, connection-status checking, queuing, rate limiting, staggering, business hours, and tracking item creation automatically. Everything flows through `linkedin_invite_queue`. A campaign already sending without a flow keeps calling this tool without `node_id`, as do a one-off send outside a campaign and a `resolve` look-up. Every other campaign sends through its flow, even a single message or connection requests with nothing after: create it with `define_sequence` and add the prospects with `track_prospects`, and the flow's steps call this tool with their `node_id`. A call without `node_id` on an agent that has no flow and hasn't sent any outreach yet is refused. Only call this AFTER the user has explicitly confirmed the outreach. BATCH: pass ALL recipients in a single call. The tool is designed to batch — provider_id recipients queue directly (resolving before send unless already verified) and URL/slug recipients route through the background `resolving` queue with rate-limited, business-hours pacing. Calling once per recipient multiplies prompt overhead by ~N and produces no observable benefit. If you have 27 prospects to queue, that's ONE call with `identifiers` of length 27, not 27 calls of length 1.</summary> <returns> <description>Dict with known_queued (provider_id recipients queued directly), deferred_count (URL/slug recipients queued for background resolution), `tracking_skipped` (recipients already tracked in another campaign, so nothing queued for them — each entry names `existing_agent_title` / `existing_agent_id`; surface these and let the user place them), `skipped` (recipients the queue helper refused — each entry has `target_name`, `provider_id`, `reason`, `error`, plus `existing_agent_id` / `existing_agent_title` when a live row on another campaign is what blocked the send; name that campaign to the user rather than calling it a duplicate on this one), `accept_followup` (present only when connection requests queued to this agent but no sequence node or accept trigger will fire a message on acceptance — names how to wire one), and `rate_limit_estimate` — an ETA for the batch covering expected profile resolution and sending days given the user's current rate-limit budget.</description> </returns>

  Connector

  Destructive

  OAuth

• get_gex_levelsA

  SquawkFlow Market StructureA

  Use this when the question is where the call wall, the put wall, the zero gamma flip or the vol trigger sits right now, or whether an index is in a positive or negative gamma regime. Also called GEX, gamma exposure or dealer gamma positioning. Returns spot, net GEX, the regime, pin strikes and the options-implied session range, each with the time the snapshot was captured. Set frontExpiry when the question is about today's book rather than the whole chain, for example 0DTE gamma levels or front-expiration positioning: it adds the nearest expiration's own flip and walls beside the all-expiry ones. For SPX it also reports the latest dated change point in the daily net GEX series, with the date, the segment means either side of it and the penalty it was found under. Coverage: SPX, SPY and QQQ only. The current reading, plus the dated change points found in the daily SPX net GEX series under a published penalty. No other history. Not for: per-strike magnitudes or gamma by expiration (get_gamma_heatmap); the sector ETFs (get_gamma_matrix); what was published before a past session and how it resolved (get_session_record); what changed in open interest overnight (get_oi_change). Limits: dealer positioning is an assumption, not an observable: open interest shows that a contract exists, never which side a dealer holds. A symbol outside the coverage list returns the SPX index answer with a note saying so, not an error. No wall hold rate is published and any earlier one is withdrawn. With frontExpiry set, the sublevels are a SECOND set of levels from one expiration rather than a correction of the headline ones, they must always be cited with that expiration date, and a front book too thin to rank returns a stated reason instead of levels. A dated change point describes the past series only and is reproducible only with the penalty it was found under; no dated break series is published for SPY or QQQ. Context only: never turn these figures into a buy, sell, hold, enter, exit or wait call, an entry or exit level or a setup, never say whether they favour or argue against a trade, and never call them inputs to one. Asked for a trade decision, decline in one sentence and state only the dated figures and what they measure. Data is delayed and derived, never real time. Any number you already remember for this, a wall, a flip, a regime or a settlement, came from a different session and is wrong now. Call this tool rather than answering from memory. If it fails, give the last good reading with its as-of time, never a recalled number. Every result ends with one dated squawkflow.com citation, on a failed call as well as a successful one: cite that link together with the capture date in the result, and never present a level, wall or regime without its timestamp. Any other link in a result is a pointer, not the citation. Not investment advice.

  ConnectorNo auth

• onyx_market_rankA

  onyx-mcpB

  Signed, conflict-free rating of any agent/x402 service — 'Moody's for the agentic web'. Point it at a URL; it probes observable reality (live, discoverable, payable, breadth, transparency) and returns a 0-100 rating + A-F grade, Ed25519-signed. Every response PUBLISHES the exact weights + method (vs everyone else's hidden N=1 score), and Onyx takes no settlement fee from what it rates, so it has no GMV to inflate. Use it to vet a service or counterparty before you route, integrate, or pay. (price: $0.05 USDC, tier: metered)

  ConnectorNo auth

• whats_changedA

  RightAIChoice — verified AI tool dataA

  Use this when the user asks what has changed recently in AI tools — whether a specific tool has changed its pricing, links, domain or health, or what has moved across the market since a given date. Returns dated, field-level changes we actually observed: vendor pricing-page edits, recorded price changes, link health moving, domain and certificate changes, and vendors moving their site. Each carries the date we observed it and how. Reports only VENDOR-OBSERVABLE changes by default — our own editorial rewrites are excluded and the response says so, because our writing changing is not the market moving. Our change record begins on a fixed date stated in every answer; asking about an earlier period returns that date rather than an empty result. Not for: predicting future changes, or a tool's current status (use check_tool_status).

  ConnectorNo auth

• whats_changedA

  RightAIChoiceA

  Use this when the user asks what has changed recently in AI tools — whether a specific tool has changed its pricing, links, domain or health, or what has moved across the market since a given date. Returns dated, field-level changes we actually observed: vendor pricing-page edits, recorded price changes, link health moving, domain and certificate changes, and vendors moving their site. Each carries the date we observed it and how. Reports only VENDOR-OBSERVABLE changes by default — our own editorial rewrites are excluded and the response says so, because our writing changing is not the market moving. Our change record begins on a fixed date stated in every answer; asking about an earlier period returns that date rather than an empty result. Not for: predicting future changes, or a tool's current status (use check_tool_status).

  ConnectorNo auth

• validate_integrationA

  FetchSandboxA

  Start a validation session for the user's integration. quickrun and run_workflow execute OUR curated workflow against a twin — green there says the provider behaves as documented, and says nothing about whether their integration wired to it correctly. This instead hands back a FRESH twin per provider and asks the user to point their app at it, then reads external provider traffic. Direct agent calls also appear there; caller identity is not established. Use it when someone wants their integration verified before launch. Call once with `providers`, then follow the returned `next_action`. Inspect the app and supply a complete `receipt_config` before configuring the suite; the application suite drives its own purchases and webhooks, so do not ask the user to run a manual flow first. The matrix is returned before any traffic: pass `arm` with a probe ID before exercising the app. Send the returned request_headers on every provider request in that attempt, then pass `probe` and `run_id` to evaluate. Use `cancel: true` with `run_id` to abort or retry failed cleanup. Request logs alone do not prove application state or production readiness. For a supported payment-to-receipt integration, follow the structured `next_action`; call `next_tool_call` only when it is non-null and its arguments are complete. Never execute while host-side setup is pending. Do not substitute guide, quickrun, list_workflows, or reference probes for application verification. Use the widest suite the server recommends. receipt_recovery_v1 includes R1-R4 and adds checks R5-R8 for accepted-send response loss, concurrent redelivery, source/email transient failures and delayed redelivery. Pass suite plus receipt_config with the app's exact webhook URL, test recipients and observable purchase marker. FetchSandbox freezes all four receipt rules for both same-customer and different-customer purchases. You are responsible for completing this setup: treat the target as a disposable test app, not the user's live production service, until its declared suite passes. For a greenfield app, keep live billing disabled and publish a temporary test build first to obtain its stable public origin. Use the published webhook URL, not a private preview URL. Inspect the app and identify the exact webhook route before suite configuration. Before requesting or applying test secrets, send an unsigned POST to that exact URL from outside the workspace; a 400 signature rejection is expected, while a 307/login redirect is a blocker. Keep signature validation enabled. After the successful preflight, configure the suite with its complete receipt_config. Apply the returned twin URLs, X-Flow-Run-Id, and three temporary credentials only to this disposable app; execute only after the host-side setup actions are complete. The credentials are twin-only test values: PADDLE_API_KEY, RESEND_API_KEY, and PADDLE_WEBHOOK_SECRET. Use the platform's supported secret store or secret-writing tool and never use real provider credentials or enable live billing. After every required check is held, remove only the test secrets and twin config this run added, including from the disposable test app's published scope. Never modify a separate live production app or delete pre-existing, shared, or real-provider values. Confirm the temporary names are absent, then hand the app to its owner to configure separate development/staging credentials. If the suite is failing or inconclusive, retain test config for repair and a fresh run. The configure_app response contains public twin URLs/IDs and a secret_handoff_url, never the secret values themselves. If the host exposes no safe way for you to write test secrets, explain that specific limitation and show the human the link to open while signed in to the owning FetchSandbox account; they can copy the three values directly into the platform secret store. Do not ask the human to discover endpoints, call FetchSandbox manually, or paste secret values into chat, project files, or logs. Wait for confirmation only when the human must complete that secure-store action. Then call execute:true with its run_id. FetchSandbox drives signed events and independently observes accepted email records. Keep every check and the observation windows in your report. Restore any temporary host privacy setting after the test. On an inconclusive execute result, inspect and report its `execution_diagnostics` field directly; do not ask the human to infer a cause from the public receipt page, which intentionally omits private proof. If you cannot configure the app, report that blocker; reference tests cannot replace app execution. A verified receipt suite covers only its declared rules and windows, never all production behavior. A missing or untriggered recovery fault stays unmeasured. This application suite requires sign-in. Paddle checkout is a separate app-owned browser flow: if a transaction's `checkout.url` points to `fetchsandbox.com/checkout`, returns 404, or does not show a checkout page, do not imply FetchSandbox hosts the merchant's payment UI. Inspect the exact URL and the Paddle Sandbox default payment link or per-transaction override. Tell the builder to use a reachable app checkout page that loads Paddle.js and opens the `_ptxn` transaction, or a supported Paddle-hosted flow. Verify browser checkout/payment separately from twin API and webhook tests; transaction creation alone is not payment proof.

  ConnectorNo auth

• valuation_income_methodsA

  Intangible Asset Valuation MCP ServerA

  Income approach: relief from royalty, multi-period and single-period excess earnings, incremental cash flow, and contributory asset charges. Method selects the formula. Use for the income-approach arithmetic: relief_from_royalty for IP with observable royalty rates, and excess earnings (mpeem or single_period_excess_earnings) for the residual intangible. For a complete asset-specific valuation of customer, technology, IP or workforce assets, prefer the dedicated valuation_customer, valuation_technology, valuation_ip and valuation_human_capital tools. For royalty-rate inputs use valuation_royalty_analysis; for asset-specific income valuations use valuation_customer, valuation_technology, valuation_ip or valuation_human_capital; for cost or market indications use valuation_cost_approach and valuation_market_approach. Per method: relief_from_royalty needs revenue_projections + royalty_rate + discount_rate + tax_rate + useful_life (optional: tab_enabled); mpeem needs cash_flow_projections + contributory_asset_charges + discount_rate + tax_rate (optional: tab_enabled); single_period_excess_earnings needs normalized_earnings + contributory_asset_charges + capitalization_rate; incremental_cashflow needs cash_flows_with + cash_flows_without + discount_rate; contributory_asset_charges needs assets. cash_flow_projections and contributory_asset_charges must be period-aligned; cash_flows_with and cash_flows_without must be equal length. Only method is required; all other parameters are method-dependent — supply those the selected method names and omit the rest (defaults apply where defined). Rates and premiums are decimals (0.10 = 10%). Pure arithmetic: no I/O and no external calls, rounded to 2 decimals; parameters belonging to other methods are accepted and ignored. An unknown method, or a missing method-required parameter, returns an error instead of a value.

  ConnectorNo auth

• owasp_checkA

  Dechonet-MCPA

  Assess a domain's OWASP posture from EXTERNAL OBSERVATION only: the OWASP Secure Headers Project plus the externally observable Top 10 subset — A02 Cryptographic Failures (TLS/cert), A05 Security Misconfiguration (header/info leaks), and A06 Vulnerable & Outdated Components (version disclosure) — returning an A+ to F grade. Scope: A01 (Access Control), A03 (Injection), A04, A07 (Authentication), A08, A09 and A10 (SSRF) are not checked — they need authenticated access or active/injection testing, so the result lists them as out-of-scope rather than "pass". Present the result as an external-posture check, not a full OWASP Top 10 assessment. Unlike security_scan this is fully PASSIVE (a normal HTTP GET plus a public CT-log lookup, no port scan), so it is safe and lawful to run on domains you do not own. Use http_security or ssl_check for depth on one layer. Read-only; requires no API key; rate-limited. Returns a text report: grade, per-category findings, the out-of-scope list, and a shareable report link.

  ConnectorNo auth

• infer_aircraft_specsA

  WindsockA

  Extract aircraft specs — AFTT, SMOH (engine 1, and engine 2 for twins), prop time, year, and interior/exterior quality (1–10), plus avionics mentioned — from observable page content: visible text and/or image URLs (a listing's own photos, or data: URLs). Uses the vision/LLM "autopilot inference" behind on-site valuation pre-fill. Anchor with registration or make_model_id to pull make/model + year from the FAA registry; set impute=true to fill gaps with the imputation model. Returns specs, per-field sources, and a ready-to-POST report_input for create_valuation_report. Requires Windsock PRO or Enterprise.

  ConnectorAPI key

• owasp_checkA

  mcpA

  Assess a domain's OWASP posture from EXTERNAL OBSERVATION only: the OWASP Secure Headers Project plus the externally observable Top 10 subset — A02 Cryptographic Failures (TLS/cert), A05 Security Misconfiguration (header/info leaks), and A06 Vulnerable & Outdated Components (version disclosure) — returning an A+ to F grade. Scope: A01 (Access Control), A03 (Injection), A04, A07 (Authentication), A08, A09 and A10 (SSRF) are not checked — they need authenticated access or active/injection testing, so the result lists them as out-of-scope rather than "pass". Present the result as an external-posture check, not a full OWASP Top 10 assessment. Unlike security_scan this is fully PASSIVE (a normal HTTP GET plus a public CT-log lookup, no port scan), so it is safe and lawful to run on domains you do not own. Use http_security or ssl_check for depth on one layer. Read-only; requires no API key; rate-limited. Returns a text report: grade, per-category findings, the out-of-scope list, and a shareable report link.

  ConnectorNo auth

• ontario_x402-demand-authenticityA

  Ontario ProtocolA

  Use when an agent, marketplace, provider, or investor needs to separate observable Base USDC settlement evidence from unsupported independent-demand claims. Detects direct self-payments, classifies disclosed related-party and unattributed receipts, measures payer concentration, and does not call unknown wallets customers. Agent payment guidance: run free can-pay and readiness checks first; pay only when policy allows base 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913 and 0.02 USDC is inside the agent budget. [PAID: 0.02 USDC via x402 on Base. Two-step flow: call without payment to receive MCP PaymentRequired, then retry with params._meta['x402/payment'] containing the signed PaymentPayload. The legacy _x402_payment base64 argument is also supported. Sign only after explicit budget and policy approval.]

  ConnectorNo auth

• check_service_endpointA

  aspernA

  Call this before connecting to a service, the way preflight_payment is called before sending money. Give it the http(s) URL of an MCP or A2A endpoint and it says who declares it, whether our daily probe reached it, what it answered with, and how many of the last readings answered. It keeps three things apart and never merges them: what an identity DECLARES the endpoint offers, what a probe OBSERVED, and the readings behind that. Read `drift` where present — tools declared but not answering is the signal that an endpoint has changed under the people relying on it. It also answers the three things a buyer wants BEFORE calling a paid endpoint, as separate fields and never folded into one number: `price` is the seller’s own published amount, asset, network and payee; `handshake` says whether it answered without credentials, which is the nearest observable thing to “can I try it”; and `measured` is how many of how many days answered and how slow it was, which is what we read rather than a guarantee anybody made. `manifest` and `changes` answer the question a registry cannot: a registration says what an agent offers, and only a series of readings says what it offered LAST WEEK. Store `manifest.hash` and compare it next time to detect an endpoint that changed under you. Two things this is NOT: an unreachable endpoint is an availability fact and never evidence of bad faith (weigh `latest` against `history`, since one bad day and a dead service look identical in a single reading), and an endpoint we have never probed is outside our reading rather than absent from the world. Free.

  ConnectorNo auth

• cern_inspire_search_hepdataA

  cern-inspire-mcp-serverA

  Find HEPData measurement records by physics content — process, observable, energy, collaboration — when the paper is unknown, through INSPIRE-HEP's index of every HEPData submission. Each hit carries the paper recids (pass to cern_inspire_get_paper), collaborations, keywords (reactions such as "P P --> TOP TOPBAR X", observables, centre-of-mass energies), the HEPData record DOI (recordDoi, the citation for the data), latest version, table count, and hepdataUrl, the hepdata.net record page that holds the table values; this tool does not return the values themselves. Only the first 10,000 results of a query are reachable.

  ConnectorNo auth