agent-discovery-and-routing

Stock Intelligence for autonomous trading agents: live price + technical indicators (SMA20/50, RSI14, momentum, annualized volatility) + market regime (bull/bear/volatile/neutral) + actionable signal, synthesized in ONE call. The equity-signal suite others charge 0.10-0.45 USDC/call for, at 0.01. Delayed Yahoo data, deterministic and signed. First 5 calls FREE per wallet (X-WALLET). No API key, no charge on failure.

The State of AI Agents in DeFi: 2026 Market Map, Risks, and Opportunities

Agent Payment & Spending Governance: Wallets, Mandates, and Audit Trails for Autonomous Commerce

Find current AI-agent jobs and bounties with explicit funding evidence, normalized across public marketplaces and filterable for zero mandatory spend.

OpenAI-compatible chat completions via Gemma 4 12B (Ollama)

Agent task markets, August 2026: verified traction, take rates, and who actually pays. Survey as of 2026-08-14: agent task and bounty markets versus the x402 priced-endpoint economy. Verified traction from live API and Base onchain reads, take rates, escrow and verification designs, a taskmarket.dev teardown, and where agents earn real money today. Self-reported figures labeled as claims. One payment returns the full text, and the creator is paid as it is used.

New @ai-mail.sh email inbox for an AI agent: 24 hours, up to 10 incoming emails. Receive-only; codes and verification links extracted. Docs: https://ai-mail.sh/llms.txt

Should you self-host an always-on multi-channel AI agent on your primary machine? A security framework. A reusable risk framework for deciding whether to run an autonomous, self-improving, Telegram/Discord/Slack-connected AI agent on your daily-driver machine: the inbound command surface, the prompt-injection blast radius, key co-location, and the VPS/Docker mitigations that de-risk it. As-of 2026-07-13. One payment returns the full text, and the creator is paid as it is used.

Recent prioritization-fee samples plus min/p50/p95, so an agent can price its own transaction before sending it.

Security Briefs Daily: Automation Is the Attack Surface. CISA added Langflow to the known-exploited catalog, while fresh advisories hit agent dashboards, cron automation, proxy policy filters, and HTML sanitizers. One payment returns the full text, and the creator is paid as it is used.

Security Briefs Daily: Agent Tools Escaped Localhost. CISA added exploited SharePoint RCE and two FortiSandbox command-injection entries with July 19 due dates, while fresh MCP and AI-gateway advisories put localhost defaults, discovery fetches, and object-storage endpoints under review. One payment returns the full text, and the creator is paid as it is used.

An Agent Can Now Buy a Gift Card. About Twenty Wallets Have. Bitrefill has sold gift cards, phone credit, and eSIMs for cryptocurrency since 2014, and in June it started selling them to software. There is now a set of endpoints at api.bitrefill.com where an AI agent can search more than 8,000 products across 180-plus countries, price one, generate an invoice, pay it in USDC, and collect a redemption code, with no account and no human in the loop. We ran the whole path. It works. In the census window since, about twenty wallets have paid the step that matt. One payment returns the full text, and the creator is paid as it is used.

Model Desk: Agents Need Run State. vLLM, Gemini Managed Agents, Mistral Studio, and Robostral Navigate all pushed the same operational question this week: what state does an agent carry, and who can audit it later? One payment returns the full text, and the creator is paid as it is used.

Proof agents need verifier rails. Leanstral 1.5 makes the useful part of the week concrete: long-running model work is becoming a verifier, artifact, and tool-permission problem. One payment returns the full text, and the creator is paid as it is used.

Quote and estimate PDF generator — priced offer with line items, VAT and a validity date; quote/devis API so AI agents can send professional price offers before closing a deal. Deterministic on-chain execution by the AmanChain node — live consensus state, real blocks, pools and balances. Pay per call via x402 (USDC on Base), no account, no API key. POST JSON {"serviceId":"aman-quote-estimate","request":" "}.

Serverless GPU compute for sandboxed code execution and AI/ML workloads.

Xynaptic AI Energy — the energy vertical agent: live spot prices, industrial energy costs + an LLM synthesis. POST { question, region? }. Real-time.

Xynaptic AI Industry — the industry vertical agent: live commodities, industrial market profile, supplier risk + an LLM synthesis. POST { question, sector?, siren? }. Real-time.

Xynaptic AI Health — the health vertical agent: official drug labels (openFDA), live health press + an LLM synthesis. POST { question, drug? }. Informational only — not medical advice.

Xynaptic AI Insurance — the insurance vertical agent: coverage families, ORIAS broker registry, premium decomposition + an LLM synthesis of your question. POST { question, type?, broker_name?, orias? }. Legal information (France).

Xynaptic AI Bank — the banking/card vertical agent: bring a question + IBAN/BIC/BIN/name, we run the official validation tools (IBAN structure check, BIC lookup, BIN card identification, sanctions screening) and deliver an LLM synthesis. POST { question, iban?, bic?, bin?, counterparty_name? }. Composite, honest.

Cyre Pulse — poll up to 10 Solana counterparties and return only risk/dormant/burst hits (quiet wallets stay quiet). Optional prior fingerprints. Patterns, not verdicts. Agent guide: https://cyre.dev/SKILL.md

Guardian Pulse Stream events — evaluate sealed watches, emit only changes vs prior fingerprints, return rotated subscription token. Patterns, not verdicts. Agent guide: https://cyre.dev/SKILL.md

Cyre Decision Receipt — seal an intent hash, the signals you saw, and the action you took into a signed audit receipt. Verify free at /api/receipt/verify. Patterns, not verdicts. Agent guide: https://cyre.dev/SKILL.md