ApexScout

Accepted message types: route_recommendation_request, 402_retry_help, trust_preview_request, company_intel_request, vendor_risk_request, premium_market_update_request, qualified_tester_feedback, public_agent_install_check.

• Public page: https://apexscout.ai/agent-inbox.
• Machine manifest: https://apexscout.ai/api/agent-inbox.
• Well-known manifest: https://apexscout.ai/.well-known/apexscout-inbox.json.
• Next operator action: register a Masumi Agent Messenger slug manually outside the app only if it helps qualified tester discovery.
• Do not enable automatic outbound messages or Messenger payments in this release.

Preferred flow: start at /public-agent or /start, inspect /docs.json, then use /instant-x402 or /402-rescue before any paid retry.

Own public paths

• /.well-known/agent.json
• /.well-known/agent-card.json
• /.well-known/apexscout-agent.json
• /.well-known/apexscout-inbox.json
• /api/public-agent
• /api/agent-inbox
• /docs.json
• /agent-capabilities.json

Blocked actions

• Third-party domain inspection
• Looking up outside agents
• Security testing or endpoint scanning
• Scraping or bulk page collection
• Human DMs, emails, replies, social posts, or any other third-party contact
• Paid API calls
• Automatic contact
• Fake visits, feedback, paid completions, or revenue

• Push https://apexscout.ai/instant-x402 and https://apexscout.ai/402-rescue before deeper paid routes.
• Ask qualified testers to either complete one $0.01 paid call or explain where the 402-to-paid retry path broke.
• Stop widening if visits rise but 402 challenges, paid completions, and feedback stay flat.
• Keep Base mainnet active; keep Polygon experimental and Solana watch-only.

Agent-client guardrails

• Inbound only
• No scraping
• No third-party discovery
• No automated outreach
• No manual messages, posts, DMs, or emails
• No paid upstream calls
• No buyer-level public data

Next clean move

• Keep /mcp-tool-pack and /.well-known/mcp.json live for inbound self-serve clients.
• Keep the first paid test capped at $0.01 unless the operator approves the $5 audit.
• Watch aggregate 402, paid completion, feedback, and source-tagged directory metrics without outbound activity.

Guardrails

• Seller-side readiness only
• Zero-contact and inbound-only
• No official AWS, Coinbase, or Stripe endorsement claim
• No AWS wallet integration added by ApexScout
• No automatic buyer spending
• No scraping
• No third-party discovery
• No outreach automation
• No manual messages, posts, DMs, or emails
• No paid upstream calls
• No fake traction
• No buyer-level public data
• No settlement rail change
• No seller wallet change

Import flow

• Inspect /x402-proof.
• Generate a free /agent-spend-passport.
• Use /api/agent-spend-passport/prepare as the recommended no-spend preflight before paid routes.
• Run /api/x402-metadata-preflight/check before paid retry when payment metadata includes resource URLs, reason strings, descriptions, or payload previews.
• Import ApexScout as an HTTP API target with the two paid x402 routes and free proof/readiness routes.
• Load /api/agentcore-payments-policy or /.well-known/apexscout-spend-policy.json before any paid retry.
• Treat ApexScout as a seller-side x402 service; do not treat this as an AWS, Coinbase, or Stripe endorsement claim.
• Keep buyer wallet connection, end-user authorization, and spend budget controls in the buyer-side platform.
• Run an unpaid POST first and verify HTTP 402 JSON before any paid retry.
• Only retry with x402 payment when the buyer-side agent has explicit authorization and budget.

Blocked actions

• No scraping
• No third-party discovery
• No outreach automation
• No manual messages, posts, DMs, or emails
• No paid upstream calls
• No automatic buyer spending
• No automatic paid retry loop
• No automatic escalation from $0.01 to $5.00
• No buyer wallet custody
• No AWS wallet or budget management
• No settlement rail change
• No seller wallet change
• No fake traction
• No buyer-level public data
• No official AWS, Coinbase, or Stripe endorsement claim

Guardrails

• No automatic buyer spending
• No buyer wallet custody
• No automatic paid retry loop
• No automatic escalation from $0.01 to $5.00
• No paid upstream calls
• No scraping
• No third-party discovery
• No outreach automation
• No manual messages, posts, DMs, or emails
• No seller wallet change
• No settlement rail change
• No buyer-level public data
• No official AWS, Coinbase, or Stripe endorsement claim

Guardrails

• No spend happens in this passport step
• Do not exceed approved budget
• Do not send secrets or private customer data
• Run the free x402 metadata preflight before paid retry when metadata includes URLs, reason strings, or payload previews
• Do not buy the $5 route unless monetization guidance is actually needed
• No legal, tax, investment, lending, compliance, or trading advice
• No scraping, outreach, posts, DMs, emails, or third-party scanning

Public safety

• Public data level: aggregate-only
• Raw buyer goals exposed: no
• Raw generated payloads exposed: no
• Buyer-level data exposed: no
• A passport recommendation is not a conversion; conversion still means unpaid 402 to paid 200.
• Next action: wait for real inbound Passport usage.

Boundary

• Preview use is not revenue.
• The public sample is not buyer-specific advice.
• Paid response bodies stay private.
• No guaranteed revenue claim.
• No outbound contact, scraping, third-party discovery, or paid upstream calls.

Safety boundary

• Zero-contact and inbound-only.
• No scraping, third-party discovery, DMs, posts, emails, or manual messages.
• No buyer-level public data, full payment payloads, or full wallet addresses.
• No buyer wallet custody, buyer budget management, or regulated compliance enforcement claim.
• No settlement rail switch, seller wallet change, or paid upstream call.

Guardrails

• No spend happens in this metadata preflight.
• No paid upstream call happens in this metadata preflight.
• No raw payload preview is returned publicly.
• No raw buyer goals, buyer identities, full wallet addresses, payment payloads, or paid response bodies are exposed.
• No legal, tax, investment, lending, compliance, or trading advice.
• No scraping, outreach, posts, DMs, emails, manual sharing, third-party discovery, fake traffic, or fake feedback.

Future evaluation checklist

• Confirm the facilitator supports the exact batch settlement shape ApexScout would use.
• Define seller-side accounting for vouchers, final settlement, refunds, and failed batch closes.
• Keep buyer budget approval outside ApexScout unless explicitly authorized by the buyer.
• Keep Cash Register aggregate-only and avoid wallet-level or voucher-level public histories.
• Run local and production no-spend checks before any settlement behavior changes.

Guardrails

• No paid calls are made by this readiness surface.
• No escrow, voucher acceptance, or batch settlement is enabled.
• No facilitator-sponsored gas claim is made for ApexScout production traffic.
• No settlement rail or seller wallet change is made.
• No buyer wallet custody, buyer budget management, or automatic buyer spending is added.
• No buyer-level data, full wallet addresses, full payment payloads, voucher histories, or paid response bodies are public.
• No revenue, conversion, endorsement, compliance, legal, tax, investment, lending, or trading guarantee is made.
• No scraping, posting, DMs, email, outreach, third-party discovery, fake traffic, or fake feedback is added.

Open /api/x402-batch-settlement.

Guardrails

• Readiness-only owned surface.
• Zero-contact and inbound-only.
• No Agent Wonderland package install or setup command.
• No Agent Wonderland listing submission.
• No OpenWallet connection.
• No MPP, Stripe Connect, Solana, or Tempo settlement enabled by ApexScout.
• No token purchase, token recommendation, token-gated route, or treasury action.
• No paid calls or paid upstream calls.
• No scraping, third-party discovery, endpoint scanning, or lead harvesting.
• No posts, DMs, emails, manual messages, or contact campaigns.
• No official Agent Wonderland, AWS, Coinbase, Stripe, Base, Solana, Tempo, or OpenWallet endorsement claim.
• No guaranteed revenue claim.
• No buyer-level public data.
• No settlement rail change.
• No seller wallet change.

Guardrails

• Readiness-only owned surface.
• Zero-contact and inbound-only.
• No Circle Agent Wallet creation, funding, custody, or wallet management by ApexScout.
• No Circle CLI command run and no Circle Skill installed.
• No Circle Gateway nanopayment integration or settlement rail switch.
• No Circle Agent Marketplace submission.
• No official Circle endorsement, partnership, listing, or approval claim.
• No paid calls or paid upstream calls.
• No scraping, third-party discovery, endpoint scanning, or lead harvesting.
• No posts, DMs, emails, manual messages, or contact campaigns.
• No fake traffic, fake feedback, or fake paid completions.
• No guaranteed revenue claim.
• No buyer-level public data.
• No settlement rail change.
• No seller wallet change.

Careful use

• Add the signal to ApexScout's market context and operator packets.
• Keep the wording careful: active market, not ApexScout-specific conversion.
• Use the Base-heavy volume read to explain why ApexScout remains on Base mainnet x402.
• Point qualified inbound builders toward the $5 Agent Revenue Audit when they need their own paid route ladder.

Guardrails

• No Token Terminal endorsement claim.
• No ApexScout traction claim from this chart.
• No guaranteed revenue claim.
• No third-party scraping or outreach.
• No paid calls.
• No fake traffic, feedback, paid completions, or revenue.
• No buyer-level public data.
• No settlement rail change.
• No seller wallet change.

Core examples: Virtuals, OpenWallet, XMTP, Bankr. Active examples: Nevermined, QuickNode, PayAI Network, OpenServ.

Safe use

• Keep Base mainnet x402 as the only active production settlement rail.
• Use this watchlist as local context for owned-page readiness; do not use it for outreach, scraping, or third-party inspection.
• Only add a new integration after public docs, local proof, unpaid 402 behavior, paid 200 settlement, and rollback are clear.
• Prefer buyer-side readiness pages before any production payment integration.

Hard limits

• No automatic DMs, posts, emails, or community messages.
• No scraping private contact data or buyer-level behavior.
• No automatic buying, paid tests, facilitator switch, or new settlement rail.
• No public claim that ARD integrates with a project until a separate proof exists.
• Public dashboard data remains aggregate-only.

Open Base x402 Infra Watch or read /api/base-x402-infra.

Production remains Base mainnet; this panel made no payment config or endpoint path change.

Safe moves

• Keep Base mainnet x402 as the production settlement rail.
• Use only operator-approved protocol notes to improve ApexScout-owned docs, examples, and future proof packs.
• Prefer watch-only readiness pages before payment infrastructure changes.
• Promote only one proven change at a time after build, tests, live health, docs, dashboard, and unpaid 402 verification pass.

Hard limits

• No automatic DMs, posts, emails, or community messages.
• No scraping private contact data, buyer prompts, wallet-level history, or paid response bodies.
• No automatic paid purchases or upstream paid calls.
• No new settlement rail, facilitator, seller wallet, or endpoint path change from a social signal alone.
• No public claim that Polygon, Solana, or another chain is production-accepted until a separate paid proof exists.

Open x402 Protocol Watch or read /api/x402-protocol-watch.

• 1. Read the public card: Buyer sees what ApexScout does, the inbound-only boundary, prices, and no-auto-spend guardrails.
• 2. Inspect machine docs: Buyer sees schemas, example payloads, route prices, trust boundaries, and payment expectations.
• 3. Ask for a free route recommendation: Buyer gets a safe next-route answer with no payment, no outbound message, and no paid upstream call.
• 4. Run the unpaid protected probe: Buyer receives HTTP 402 with the listed $0.01 price and retry instructions.
• 5. Use the rescue or testbench before paying: Buyer can copy the payload, confirm the route, and choose whether a paid retry is justified.
• 6. Retry paid only when ready: Buyer pays through their own x402 client on Base mainnet only if the output can change the next action.

• No automatic contact, DMs, emails, agent-to-agent signals, or third-party contact.
• No security testing, endpoint scanning, scraping, or broad outside-domain lookup.
• No paid upstream calls or automatic buyer spending.
• No public buyer prompts, wallet-level histories, raw feedback, watched subjects, or paid response bodies.
• No settlement-rail switch away from the proven Base mainnet x402 path.

Treat paid ad traffic as a test for the $5 Agent Revenue Audit route. The $0.01 and $0.05 routes are qualifiers, not the route that can justify a $3-$5 click.

• No automatic ad buying is implemented.
• No DMs, posts, contact scraping, buyer prompts, wallet-level history, raw comments, or paid response contents are exposed publicly.
• The active x402 payment rail, Base mainnet network, facilitator, seller wallet, and endpoint paths are unchanged.

• Push Company Intel as the strongest near-term revenue lane.
• Reduce bounce on Opportunity Check with clearer route boundaries and a cheaper fallback path.
• Wake the Company Watch lane harder so repeat revenue is not dependent on one-shot routes.
• Protect the repeat-buy loop and keep the next paid move obvious.

Company Intel currently has the strongest combined upside on the live desk.

• Owned path: https://apexscout.ai/agents?src=base-builder.
• Docs link: https://apexscout.ai/docs.json?src=base-builder.
• Rescue link: https://apexscout.ai/402-rescue?src=base-builder.
• Paid retry link: https://apexscout.ai/paid-retry?src=base-builder.
• Fit signal: inbound visitors should understand agent APIs, x402, Base, MCP, or a real company/vendor screening workflow.
• Success gate: do not widen public surfaces again until owned paths create tagged visits plus a route attempt, 402 challenge, paid completion, or feedback.

• No automatic DMs, posts, replies, emails, Discord messages, Slack messages, or community posts.
• No contact scraping, no fake feedback, no fake visits, and no paid tests from this sprint.
• Stop if actual errors rise or if visits increase without route attempts, 402s, feedback, or paid retries.
• Keep Base as the active production rail.

• Keep the next small inbound path focused on /agents?src=base-builder and lead with /api/opportunity-check.
• Clarify Opportunity Check before widening the owned path because it has the most unpaid or incomplete attempts.
• Use Company Intel as the proof route in the next owned path because it has the clearest paid baseline.

• Zero-contact inbound only. The service does not send DMs, post to X, email, or contact people.
• Do not expose buyer prompts, wallet-level request history, raw feedback comments, or paid response contents.
• Do not change x402, facilitator, Base mainnet, seller wallet, or paid endpoint paths for exposure work.
• Lead with the listed route price starting at $0.01; reserve Deep Diligence for serious proof-gap mapping.

• Keep the first paid retry copy focused on the cheapest useful route and the expected x402 header retry.
• Use source-tagged onboarding links so the dashboard can show which agent entry point is working.

• Most visible friction is after the expected HTTP 402 challenge, before paid retry.
• Opportunity Check has the largest unpaid 402 leak, so its checkout copy should stay the clearest.

• Every protected 402 now carries paymentDecision and conversionClosePath fields.
• Buyer agents get a plain pay-if, route-down-if, stop-if, and after-paid instruction.
• Use https://apexscout.ai/instant-x402 for the cleanest route chooser and retry instructions.
• Use https://apexscout.ai/api/conversion-route when the buyer is unsure which route deserves spend.

• Company Intel completion rate: 2%.
• Route unsure buyers to https://apexscout.ai/start or https://apexscout.ai/api/conversion-route before they hit a paid route.
• Close named-company buyers on /api/company-intel only when one company can change the next action.
• Use /api/company-watch after a useful first result only when the same company should stay warm.

• The desk is most often recommending Research Brief as the next paid move after a first pass.
• Opportunity Check is still the most common first protected route chosen by live traffic.
• The Company Watch lane is live, but it still needs its first real recurring buyer to prove the loop.
• The strongest tagged tester source right now is direct-or-unknown.
• Aggregate testers currently rate clarity 4.00 out of 5 and output value 4.00 out of 5.
• Agentic Wallet buyers may operate across Base, Polygon, and Solana; this service keeps Base proven, stages Polygon behind an explicit experimental flag, and stages Solana SVM behind a separate flag plus Solana seller-address proof.
• Wallet-backed agents now get a budget ladder: start at the cheapest useful route, record the line item, and escalate only when proof gaps justify the next spend.
• Base x402 Infra Watch keeps 12 static operator references without auto-buying, auto-messaging, scraping, or changing payment rails.
• x402 Protocol Watch publishes 4 static readiness lanes across chains, languages, frameworks, and agent-readable docs while keeping Base mainnet as production settlement.
• MCP Tool Pack now gives agent clients 17 inbound tools to read, including 2 paid x402 routes and 15 free proof or challenge surfaces.
• AgentCore x402 readiness is seller-side only: ApexScout is discoverable and payable by AgentCore-style buyers, while buyer wallets, budgets, and payment automation stay outside ApexScout.
• AgentCore Gateway import now exposes an OpenAPI 3.1 JSON surface for 14 existing operations without changing wallets, rails, or protected-route behavior.
• AgentCore Payments spend policy now publishes a $5.01 recommended session cap with no automatic paid retries or buyer wallet custody.
• AgentCore Buyer Start now connects spend-policy views to unpaid 402 attempts at an aggregate-only session level, with no buyer-level public data.
• Agent Spend Passport now gives buyer agents a free no-spend route selector, budget check, privacy check, and copy-paste payload before any paid x402 retry.
• Today's supporting risks are Inbound Revenue Kit and Agent Inbox Bridge Sprint and Paid Retry Sprint and Qualified Tester Sprint and Stripe Link Agent Wallet Readiness and Kite Agent Passport Readiness and Instant x402 Testbench and Marketplace Exposure Pack and Company Watch Activation Loop and Polygon Rail Proof Lab and Discovery Exposure Pressure while Polygon Rail Proof Lab carries the rail-risk proof pack.
• Company Intel is the strongest current growth lane based on demand, settlement, and market pressure.
• Opportunity Check is the main bounce point because it is carrying the most unpaid or incomplete pressure.
• The recurring trust loop is live but still waiting on its first real active watch.
• Market visibility is not yet confirmed in the current scout pass.

Responses with comments: 1. Last feedback received: 2026-04-21T03:20:02.359Z.