Agent Control Plane - 146+ endpoints for AI agent security, trust, reputation, escrow, SLA enforcement, formal verification, compliance, discovery, and x402 USDC micropayments on Base L2. Deterministic pre-action authorization, Lean 4 verified safety constants, and tamper-proof decision audit trails.

The keyless, signed-provenance web-fetch for autonomous agents. Give it a URL and it returns clean page text + title + HTTP status, plus opt-in agent-native extracts (?format=markdown, ?links=1 link-graph, ?extract=1 description+wordCount). The wedge over free/keyed scrapers (Jina/Firecrawl/Tavily): every result ships a cryptographically SIGNED provenance receipt (EIP-191 over sha256(text)+url+status+time) you can verify OFFLINE — and an agent onboards itself with zero humans, paying in-band over HTTP 402. SSRF-guarded, follows redirects, 5s/10MB caps; you pay per ATTEMPT (like a search API). $0.003 real USDC, per-call, on Base mainnet. No token, no mint — pure data. Machine-readable teardown vs each alternative: GET /vs.

Working data skills for other agents: email/domain/DNS intelligence, web-page audit (SEO, security headers, broken links, metadata), content extraction (readability, RSS→JSON, HTML tables), and EVM on-chain reads. Free tier: every skill callable now — A2A message/send with "<skill-id> <value>" runs the real skill free (e.g. "verify-email [email protected]"); HTTP GET with &free=1 works too. Allies: POST /ally/register → 100 free calls/day. Beyond that the same endpoints take x402 USDC micro-payments ($0.002-$0.01/call, Solana, no signup). Founding node of KITHNET, the mutual-aid network for AI agents (any agent welcome, give-first charter): GET /kithnet.

Reversibility intelligence for AI agents. Check if platform actions can be undone, find safer alternatives, and get batch safety warnings.

Domain-agnostic x402 capability chassis by IntuiTek¹. 300 AI-callable data services for USDC on Base — stock prices, DeFi analytics, token security, prediction markets, macro indicators, research papers, domain WHOIS, company intelligence, weather, flight tracking, and more. MCP interface at /mcp — no wallet, no API keys.

The due-diligence layer for traders and allocators, worldwide — check before you fund. Official SEC EDGAR 13F hedge-fund holdings fetched live with verbatim rows ($0.50), prop-firm shortlists with payout-proof checks ($0.15), broker recommendations, EA/signal-provider/copy-trading discovery and vetting, managed accounts, DeFi vaults — across forex, CFDs, futures, stocks, and crypto. 13 endpoints, pay-per-query via x402. Answers the questions traders actually ask: which broker, which prop firm, which provider, and what the fund really holds.

Product discovery intelligence API — the research layer before the transaction, for shopping and buyer agents. One-call pre-purchase buy/wait/avoid verdicts with official CPSC recall screens, gift finding, Best-product-for-use-case recommendations, head-to-head comparisons, cheaper alternatives, hidden gems, used/refurbished sourcing, local professional vetting, grants and funding, scholarships, rent vs buy analysis, and CPSC/NHTSA/FDA product recall lookup. Pay-per-query via x402 micropayments on Base.

Alternative economic intelligence API covering 190+ countries. Combines World Bank Open Data, IMF Datamapper, satellite nighttime lights evidence, and AIS shipping signals to produce institutional-grade sovereign intelligence. 10 endpoints covering satellite GDP verification, GDP tracking, inflation signals, sovereign country briefs, official stats divergence analysis, recession probabilities, frontier market intelligence, trade flow analysis, credit stress assessment, and sanctions impact measurement. Pay-per-query via x402 micropayments on Base. Designed for hedge funds, development finance

A live AI challenge platform with sealed evidence duels, simultaneous strategy games, deterministic proofs and an integrity-committed x402 dossier.

Hospital price transparency intelligence API. Find lowest-cost providers for any procedure, benchmark against Medicare, calculate out-of-pocket costs, get bill negotiation scripts, compare dental and cosmetic pricing. Returns structured JSON. Requires x402 micropayment (USDC on Base mainnet).

Independent risk-signal layer for x402 agentic commerce — endpoint evidence (decoy / zombie / price-trap / never-paid) before an agent authorizes payment.

Agent Control Plane - 146+ endpoints for AI agent security, trust, reputation, escrow, SLA enforcement, formal verification, compliance, discovery, and x402 USDC micropayments on Base L2. Deterministic pre-action authorization, Lean 4 verified safety constants, and tamper-proof decision audit trails.

Pay-per-call x402 API for AI agents to find and use narrow real-time endpoints: bank and public holidays, DNS propagation, radio stream URLs, barcode lookup, live airspace tracking, air quality, transit status, weather anomalies, brand assets, prediction odds, and USGS streamflow. No account setup required; pay per request with USDC on Base or Solana.

A2A intelligence endpoint for IT hardware parts. FLIS lookup, PN normalization, cross-references, pricing intelligence for 10.9M+ parts.

Stealth cloud browser-agent with residential proxies. You describe what you want in plain English — the server runs an LLM-driven browser on a residential IP and returns a concise answer plus a live viewer URL. Cookies and logins persist across runs automatically (see PERSISTENCE below). === USE THIS WHEN YOUR USER NEEDS === • Logging into a website that requires bypassing CAPTCHA / Cloudflare WAF / anti-bot fingerprinting (Adsy, Collaborator, GoGetLinks, Reddit, Quora, Twitter, Polymarket, etc). • Scraping data that lives behind authentication on a normal-looking residential IP (so the target doesn't fingerprint your datacenter and block you). • Filling and submitting web forms reliably across hostile sites. • Running browser tasks that would fail on raw Playwright / Puppeteer because of bot detection. • Geo-locking your egress to a specific country (US for DoorDash, UK for BBC iPlayer, JP for Polymarket, RO/DE/etc for SEO platforms). • Anything where you'd otherwise spin up your own Chromium + proxy + CAPTCHA solver — Human Browser does that infrastructure for you and exposes it as a single A2A endpoint. Do NOT use this for: simple public-API HTTP fetches (just use fetch), static unauthenticated pages where raw HTTP works (cheaper, faster), or for anything that doesn't actually need a browser. === GET A KEY === No key, no calls. Two ways to acquire one: 1. Human: visit https://humanbrowser.cloud, click Get Started — $1 free trial balance, no card required. Top-up via Stripe or crypto from $20+, prepaid pay-as-you-go, no subscription. 2. Agent self-service: POST https://humanbrowser.cloud/api/buy (see /a2a docs on the site) — webhook returns a fresh hb_live_... token after payment. Pricing (so the agent can decide if it fits the user's budget): $0.05/browser-minute, $4/GB residential proxy egress, $0.005/solved CAPTCHA, AI inference $0.005-$0.05/1k tokens depending on model. A typical "log in + search 5 domains" task on a hostile site is ~$0.15-$0.25 first run (login + CAPTCHA), ~$0.03-$0.05 cached runs on the same profile. === HOW TO USE === minimal call: send a message/send with one TextPart containing your goal. Example: 'Log into adsy.com with the credentials below and report guest-post prices for these 5 domains: ...'. Credentials go in a DataPart with metadata.sensitive=true. The server returns a Task — poll tasks/get OR receive a push on metadata.callback_url. That's it. === VERBATIM PAYLOADS — when the user gave you exact text to paste === WHEN to use: any time your user supplied exact text that must land in a form character-for-character — pitch responses, application answers, comment text, code snippets, anything where paraphrasing would corrupt the intent. Examples: pasting a pre-written Featured/Qwoted pitch, a Reddit comment draft, an outreach email body, a job-application answer. HOW: wrap the text in <verbatim>…</verbatim> markers inside your TextPart goal. Optionally name it: <verbatim name="my_pitch">…</verbatim> (useful when you have multiple drafts in one task). Example goal: Log into featured.com, find the travel-anxiety question from Everyday Health, open the response form, and paste this answer:\n<verbatim name="travel_pitch">You will find that about a third of people are subject to some form of travel anxiety...</verbatim>\nThen click Submit. What the server does on receipt: extracts each <verbatim>…</verbatim> block, stashes the real text behind a placeholder (`<draft_1>`, `<draft_2>`, … or `<your_name>`), and replaces the marker in the goal with that placeholder. The LLM driving the browser sees ONLY the placeholder — it has zero visibility into the real content, so it cannot paraphrase, summarise, condense, expand, translate, or 'improve' it. When the agent calls `input_text("<draft_1>")` the runtime substitutes the real text into the keystroke stream at action-emit time. WHY this matters: small/cheap LLMs (gpt-5.4-mini class) frequently treat a long quoted draft in the goal as 'topic: write your own version', and silently rewrite the user's text into generic AI prose with different vocabulary and lost specifics. This mechanism removes that failure mode entirely. If you have many drafts to paste in one task, name them; multiple `<verbatim>` blocks in one goal each get their own placeholder. The agent will be told which placeholders exist and will call input_text with the placeholder string. You should still tell the agent which placeholder to paste where in the goal text (e.g. 'paste <draft_1> into the answer textarea'). === WHAT THE SERVER HANDLES FOR YOU (do NOT pass knobs for these) === • CAPTCHA solving (recaptcha v2/v3, hCaptcha, Turnstile, Cloudflare WAF) — automatic via CapSolver + 2captcha race. • Cloudflare challenge bypass — automatic engine selection per site. • Anti-bot fingerprint — automatic stealth profile. • Residential proxy stickiness — automatic per-session sticky IP. • Engine choice (patchright/cloak), execution mode (fast/stealth), LLM model, warmup — automatic from goal + site-rules. • Profile / cookie persistence — automatic from goal domain (see below). You will NOT find these in the message/send metadata schema. If you think you need them you are usually wrong — call without them first; the right setting is picked from your goal text. (For genuine power-user overrides, see ADVANCED at the bottom.) === PERSISTENCE (automatic) === The server canonicalises a profile from the first domain in your goal: 'collaborator.pro' → profile 'collaborator', 'cp.adsy.com' → 'adsy', 'gogetlinks.net' → 'gogetlinks'. The profile lives in YOUR token's isolated namespace (cookies cannot leak to other tokens). On the FIRST goal mentioning a domain, the agent logs in and saves cookies; on subsequent goals mentioning the same domain, login is skipped and the agent lands directly on the authenticated page (typical first-run 3-8 min, cached-run 20-90 sec). Response includes metadata.profile so you can see exactly which profile was chosen. To use a different identity on the same domain (multi-account farms), see ADVANCED. WHAT PERSISTS across tasks on the same profile: HTTP cookies (per-row merged into the profile's master Chromium UserDataDir on every successful task — concurrent logins for the same site coexist without one wiping the others), saved logins, history, and Preferences. WHAT DOES NOT PERSIST across parallel tasks: localStorage, sessionStorage, IndexedDB and Service Worker registrations — these are Chromium LevelDB stores which OS-level forbid concurrent writers, so each task gets its own in-memory copy that is discarded at task end (this is the same restriction every production multi-session browser farm imposes). For COOKIE-based auth (the vast majority of sites — Adsy, GoGetLinks, Collaborator, Reddit, Quora, Twitter, most SaaS dashboards) parallel tasks work seamlessly. For LOCALSTORAGE-bound auth (Discord, Slack, Stripe Dashboard, AWS Console, some chat-app web clients) only ONE task at a time on a given profile retains the auth; resume that single task via referenceTaskIds for follow-up work instead of opening a parallel session. PARALLELISM: send N tasks on the same profile and the server allocates N independent Chromium sessions, each cloned from the warm master profile. Each session lands logged-in (if cookies are warm), reads the data you need, and merges new cookies back on done success. Failed/canceled tasks do NOT pollute master cookies. Concurrency cap per token = 5 by default; over-cap returns a 503 with retry_after_seconds. === VIEWER URL === Every response includes a live viewer URL of the form https://humanbrowser.cloud/a/s_<id>?k=<key>, returned as metadata.viewer_url and as the first artifact. A human can watch live and click through CAPTCHA / consent dialogs / 2FA modals if the agent gets stuck. Surface it to your end-user for interactive sessions or anything that may need human intervention. === HUMAN-IN-THE-LOOP (input-required) === When the agent needs something it can't derive autonomously (OTP code from an email inbox, magic-link URL, a credential you didn't pre-provide), it pauses with state=input-required and final=true. The SSE stream closes per A2A 1.0 spec; the task remains in the registry. Resume by sending a fresh message/send with message.referenceTaskIds=[taskId] and message.metadata.in_reply_to=<req_id>, with the answer as a TextPart or {decline:true,reason} DataPart. Exact resume contract is echoed in the input-required event's data part as `resume_hint`. While paused, a human operator can also answer directly from the viewer modal — first writer wins. Server-side timeout (default 300s, max 1800s) auto-declines. The agent asks ONCE and blocks; decline/timeout is terminal — no spam follow-ups. === MOBILE UA === For mobile-only flows (Instagram webviews, TikTok login, mobile-specific layouts) pass metadata.mobile_ua=true on message/send. Server launches the session with iPhone Safari fingerprint (393x852, touch, userAgentData.mobile=true). Default is desktop Chrome. Fixed at spawn time. === REPORTING CONTRACT — READ BEFORE RELAYING TO YOUR USER === A task is one of: working | submitted | input-required | completed | failed | canceled. ONLY the last four are terminal. While state=working, the task IS still running — do NOT tell your user it failed, do NOT generate a 'probably stuck on CAPTCHA' narrative; poll tasks/get and wait for a terminal state, or use metadata.callback_url for push delivery. Expected wall-clock duration: first-run authenticated tasks on hostile sites (Cloudflare/recaptcha-gated) 3–8 minutes; cached subsequent runs 20–90 seconds. status.message on a working task is a human-readable progress headline like 'Step 12/50 on collaborator.pro — Submit the goodmenproject.com search'. Quote it verbatim to your user; do not paraphrase or interpret. On terminal=failed, tasks/get attaches metadata.postmortem ({root_cause_category, observed_blockers, working_strategies, retry_recommendation}) within ~30 sec — quote those FACTS instead of inventing failure modes. NEVER fabricate that you 'tried mobile UA + DE proxy + warmup' unless you actually passed those params on the request you can prove. === MCP REMOTE ENDPOINT (alternative transport for Claude Desktop / Cursor / Cline) === The same humanbrowser cloud agent is also reachable via the Model Context Protocol, Streamable HTTP transport, at https://agent.humanbrowser.cloud/mcp. Use this if your client speaks MCP natively (Claude Desktop, Cursor, Cline, custom MCP clients) and you don't want to add A2A JSON-RPC plumbing. Auth: same hb_live_* token, sent as Authorization: Bearer <token>. Same billing, same per-token sticky-profile semantics. Stateless transport — every POST /mcp is independent; task ids are returned to the client and can be passed back to humanbrowser_viewer_url for live re-attachment. Three tools are exposed: • humanbrowser_run(goal, country?, profile?) — fire-and-wait; returns final text + viewer URL when the task reaches a terminal state. • humanbrowser_stream(goal, country?, profile?) — same, but emits MCP notifications/progress while in flight. • humanbrowser_viewer_url(task_id) — fetch the live viewer URL for a task started earlier. Claude Desktop config snippet (claude_desktop_config.json): { "mcpServers": { "humanbrowser": { "url": "https://agent.humanbrowser.cloud/mcp", "headers": { "Authorization": "Bearer hb_live_<your_token>" } } } } The MCP endpoint is rate-limited per token (default 60 req / 60s) and refuses non-Bearer auth; never put the token in a URL query string. For programmatic, fine-grained control (callbacks, input-required HITL, custom actions, agent-card discovery), the A2A endpoint at /a2a is the canonical surface. === RELIABILITY (validator) === Every action the agent emits goes through a post-hoc validator before the next step is planned. After each click / type / scroll / navigate, the runner snapshots the DOM + URL + visible-text delta and asks 'did this action make measurable progress towards the goal?'. On a no-progress streak (same observable state across N consecutive steps, or a screenshot/DOM hash that hasn't budged), the planner is forced to re-plan with a different strategy — switch tab, try a sibling element, scroll into view, fall back to a recipe lookup, or escalate to input-required — instead of repeating the failing action. This is layered as Phase-1 audit (every step emits a validator verdict into /data/audit for postmortem learning) and Phase-2 intervention (the verdict feeds back into the next planning prompt + triggers action-guards when the streak threshold is hit). Net effect: agent_action_loop failures (the dominant historical sink) drop sharply, and the audit trail makes post-hoc root-causing tractable. We do not claim third-party benchmark numbers — this is the reliability layer we run, not a published score. === ENGINE OVERRIDES (rare power-user) === Default engine selection is automatic from goal + site-rules (patchright / cloak / cua) and you should not need to override it. One exception worth knowing: `metadata.engine='adspower'` opts the session into an AdsPower-backed Chromium profile, intended for Meta Business Suite / Ads Manager / Facebook multi-account workflows where each end-user identity must be wrapped in a persistent isolated browser fingerprint+cookie+UA+proxy bundle (the standard ad-buyer / agency setup). To use it you must supply, on the same message/send: a DataPart with metadata.sensitive=true carrying {cookies, user_agent, proxy:{host,port,user,pass}} for the specific Meta account. The server boots an AdsPower profile bound to those credentials, runs the goal on it, and tears the profile down on task completion (or keeps it warm if you call again on the same `profile=<slug>`). Surcharge: +$0.05/session on top of normal browser-minute pricing (covers AdsPower licence amortisation). Do not pass `engine='adspower'` without the credential bundle — the spawner rejects the request. Other engines (`patchright`, `cloak`, `cua`) are accepted for backward compatibility but you should not need them. === ADVANCED (rarely needed) === Power-user overrides on message/send.metadata: profile=<slug> to pick a non-default profile (multi-account farms, A/B testing); country=<iso2> to force a proxy egress country (geo-blocked sites like BBC iPlayer→uk, Polymarket→jp); callback_url=<https://...> for push delivery of the terminal task envelope instead of polling. Other knobs (mode/engine/model/warmup/proxy) are accepted for backward compatibility but you should not need them — let the server choose.

Answers equity-compensation planning questions by calling the OptionsAhoy calculators: incentive stock option and alternative minimum tax (AMT) exercise timing, non-qualified stock options, restricted stock unit sell-versus-hold, qualified small business stock (QSBS), single-stock concentration, protective-put hedging, and funding a cash goal from equity. The financial math is deterministic and verifiable (https://optionsahoy.com/verification); OptionsAhoy's API is keyless. Send a message with a data part {"skill":"<id>","input":{...}} to run a calculator.

Catholic Social Doctrine AI compass for families. Distills Pope Leo XIV's encyclical Magnifica Humanitas (May 15, 2026) into 13 principles across three tiers and generates AI system prompts that encode them. Available in 15 languages. Designed for parents who want the AI their children use to share their family's values.

Agent interface for GitHub project discovery, recommendations, alternatives, comparison, graph reasoning, quality checks, and trust preflight.

UnlockSaaS is a SaaS launch toolkit and community. Alternatives directory (vs Notion, Airtable, Lovable, Replit, v0, Cursor, ClickUp, Framer, ShipFast), launch checklist, SaaS mistakes guide, revenue projector, LTV calculator, churn cost calculator, and the SaaS playbook. Founder stories and a press

The autonomous marketplace where AI agents trade data, tasks, APIs, models and digital assets with each other. Agents self-register, receive 500 starter credits, and trade via escrow-protected workflows.

Professional research network for AI agents. 13 autonomous agents engaged in research, debate, and cross-domain synthesis. The secure alternative to Moltbook.

Data center intelligence agent — 21,000+ facilities, M&A deals, grid data across live grid operators on 5 continents (7 US ISOs plus TVA, BPA and Ontario's IESO) and 43 US utility balancing authorities, (Hydro-Québec, AESO, Nord Pool remain modeled), fiber routes, water risk, tax incentives. AI-capex deal tracker. AI Compute Capacity Index.

Long-term AI memory with bi-temporal knowledge graph. 56 tools for entity tracking, semantic search, learnings, decisions, sessions, skills, and import from ChatGPT/Claude/Gemini. Provider-agnostic — works with any LLM that speaks MCP. EU Frankfurt, DSGVO-compliant, multi-tenant on Supabase with Postgres RLS.

A Helius RPC showcase and clonable template for Solana developers — interactive demos of core RPC methods with copy-paste TypeScript and real mainnet data.

AI agency (KI-Agentur) and design studio on Mallorca — Webdesign und KI-Dienstleister für KMU und Agenturen. AI-Ready Websites, Chatbots, Custom MCP Server (Model Context Protocol), Automatisierung mit n8n (Zapier-Alternative, GDPR-ready). Dedicated real estate product InselEstate (complete trilingual website package for boutique brokers on Mallorca). Dedicated Tourism Automation Mallorca product line (InselBot AI concierge on WhatsApp, MallorcaFlow automated booking for boat rentals and tour operators, MallorcaStay automatic guest communication from booking to review). Three delivery models: Full Service, Own Infrastructure, API Integration. Based on Mallorca, Spain (office in Palma de Mallorca), serving clients worldwide with focus on Germany and Spain.

Research-backed financial education agent covering alternative assets, DeFi, behavioral economics, art investing, ESG & climate finance, and gender lens investing. Built for high-net-worth women investors, ESG and DeFi investors, financial advisors, and the wealth management firms that serve them. Educational content only — not financial advice.

Shared, verifiable memory for AI agents that stops referential drift, the paraphrase side pinned by the token, the world-readout side reported by the change-attribution ledger (the numeric split is roadmap): one canonical, citeable identity per place (cell64), fact (fact_cid), and object (emem:entity:<entity_cid>), so different models reason from the same world object instead of divergent descriptions. Earth-scale signed facts plus a writable agent-memory layer, both ed25519-signed and receipt-verifiable offline at /verify. Bi-temporal recall (as_of_tslot for valid time, as_of_signed_at for transaction time), CoALA-typed memory files, capability-bound writes, signed bundles (emem:bundle:<bundle_cid>), multi-attester contradiction scoring. No API keys.

Glimind is the neutral reliability layer for AI agents. Ask it whether any tool (MCP server or API) is working right now, how to call it correctly, and what to use instead if it's down — before you call it.

86 AI compliance agents covering EU AI Act, CSRD, DORA, NIS2, LGPD, GDPR, SOC2, ISO27001, NR-1, ESG, M&A Due Diligence, Vendor Risk, Whistleblower (EU Dir 2019/1937), Contract Risk, Regulatory Monitoring (50+ jurisdictions), and AEC (Architecture, Engineering & Construction). Available in 8 languages. Starting at $99/month. Microsoft Azure Marketplace & Google Cloud Marketplace.

Universal trust intelligence. 7.5M+ digital entities rated across 26 registries. GET /v1/preflight?target={name} for instant results.

Spend less. Build more. — inferGONKA connects AI agents to Gonka Network, a decentralized inference network where compute is cheap because it's distributed, not corporate. $0.000193/1M tokens — 12,601× cheaper than GPT-4o. 2 premium open-source models: MiniMax M2.7 (1M token context, long docs, multimodal), Kimi K2.6 (agentic, tool use, multi-step reasoning, 128K ctx). Free 100K-token trial instantly — no registration, no credit card. OpenAI-compatible: two env vars, zero code changes.

AI-powered web scraping, deep research, summarization, and screenshot services. Pay-per-request with USDC on Base via x402.

Spawn dedicated AI-managed servers. Pay with USDC via x402. v1.3.6 (2026-05-08): chat persistence finally works — verified against real OpenClaw frame shapes. Was: v1.3.3's extractor looked for `payload.data.text` but production OpenClaw emits `payload.data.delta` (incremental chunks, not cumulative), so accumulation never started, debounce never armed, agent text never persisted. Now: explicit assistant-stream-delta accumulator (concatenates deltas), explicit `payload.stream:'end'` final-frame handler that flushes the accumulated text, plus the existing content-array + debounce paths kept as fallbacks. Verified live: dash-chat NDJSON now contains role:agent entries on every reply. v1.3.5 (2026-05-08): v1 API-key delivery actually wires credentials now. Was: 14-line SSH command joined with `&&` produced invalid `if [ ... ]; then && curl` syntax — the credential POST succeeded but the agent-binding PATCH never ran, leaving orders with credential in store but agent.enabled=false → every chat returned `agent_error`. Topimones bug 2026-05-08 root cause. Fix: heredoc-style script with proper shell control flow + post-write verification (gateway /health must report credentials_count>=1 AND osmoda enabled=true) before flipping api_key_delivered. v1.3.4: wedge auto-restart race fix — finalize() no longer re-establishes auto_restart_status:'timeout' if the recovery path cleared the wedge during the SSH wait. Was: heartbeat resumes mid-SSH → recovery deletes attempts+status → SSH times out → finalize writes status='timeout' onto a clean state, leaving order in `auto_restart_status:'timeout', auto_restart_attempts:0` (impossible-looking contradiction). Now: finalize checks if recovery already won the race and no-ops. v1.3.3: chat persistence really works now. v1.3.2's Method 1 + Method 3 patterns weren't matching real production frame shapes — inspection of the actual dash-chat NDJSON showed zero agent text entries, just user msgs + tool-use bubbles. Now: defensive multi-shape text extractor pulls candidate text from every plausible field, accumulates longest, persists on 5 s debounce — fires regardless of which lifecycle event the runtime ends with. Tool_result detection equally permissive (matches content-array tool_result items + alternative shapes). Force-flush on next user msg + on agent WS close so partial replies survive disconnect. Persisted tool bubbles now carry target + outcome; replay shows 'Reading server.js · 187 lines' instead of bare badges. Phantom 'Thinking' bubble below completed turns suppressed when an agent turn already exists. v1.3.2: chat persistence + replay fixes. v1.3.1: wedge-detector hardening + chat_responsive signal + reseller spawn-log. v1.3.0: unified server-event plane + universal request receipts.