Agent-net Open Sources Webagent: A Go Harness That Turns Any Website into a Guarded AI Agent
Agent-net, the team building an agent-to-agent marketplace where AI agents discover, trust, and pay each other, has released Webagent, an open source harness for standing up public-facing business agents.So, basically you give it your website, get an agent, and let it talk to other agents.
Instead of writing orchestration code, a business fills in a declarative JSON spec, picks 1 provider for each of 9 pluggable slots, and runs webagent serve.Is it deployable?Yes, with caveats.The repo ships under Apache 2.
0, builds green, and runs live Slack, WhatsApp, and HTTP agents backed by MCP tools today.It is still labeled v0, and the browser action provider, OAuth-gated MCP, OTel export, and the AgentNet identity and billing layer are listed as not yet built.Slots, providers, picks Webagent is written in Go.
An agent is 1 Brain (an LLM plus instruction) over a set of slots defined in core/.
Each slot is a Service Provider Interface with a registry of providers in spi/ and a designated default: SlotBuilt-in providersDefaultRetrievallive, keyword, hybridliveMemorysessionsessionGuardrailbasic, offbasicChannela2a, web, slack, whatsapp (telegram stub)a2aSecretsenv, file, staticenvPresentertext, terminal (QR), webtextModelecho, openrouter, gatewayechoActionnone, demo, mcpnoneObservabilitynone, log, memorynone The design serves 3 audiences on 1 contract: businesses that configure by picking from the menu, businesses that extend by registering a custom provider, and partner companies that ship adapters.
Nobody forks the core.Every provider must pass its slot’s conformance suite to be certified.Safety is code-enforced, not prompt-enforced The most important architectural decision is action.Guard.
Every tool the agent holds, whether it comes from the action provider or is injected by the host, is wrapped so the chosen guardrail runs on the action before it executes.The model cannot bypass it.The DESIGN.
md frames the whole project around a research finding that architecture, not model capability, decides agent success, citing arXiv 2511.19477 and an 85% versus 50% task-success gap on the same models.
What runs today The default echo brain needs no credentials, so webagent validate and webagent serve work out of the box on the 2 example specs, zomato.json and bakery.json.To drive a real model, webagent keys set openrouter stores a key in the OS config directory with mode 0600.
Keys are never written into a spec, and an exported environment variable always wins.Both openrouter and gateway are OpenAI-compatible clients with a tool-calling loop.A business with an existing MCP server becomes an acting agent with 1 spec block.
The mcp action provider connects over Streamable HTTP (JSON and SSE) with bearer or API-key auth, performs the handshake at build time, and hands each tool to the agent behind the guard, so validate reports the real tool count.
The Slack and WhatsApp channel adapters verify every inbound webhook signature, acknowledge immediately, reply through the platform API, ignore their own messages, and de-duplicate retried deliveries.Slack points at /slack/events; Meta’s callback goes to /whatsapp/webhook.
Secrets follow a naming rule: any config key ending in Secret is a reference resolved through the selected vault at build time, so specs are safe to commit.A reference that cannot be resolved fails the build rather than starting a channel without a credential.
Observability emits a per-turn TurnTrace aligned to the OpenTelemetry GenAI conventions, and an eval/ harness runs scenarios with checks.window.addEventListener("message",function(e){ if(e.data&&e.data.mtpWebagentHeight){var f=document.getElementById("mtp-webagent-frame");if(f)f.style.height=e.data.
mtpWebagentHeight+"px";} }); Key Takeaways 1 JSON spec, 9 slots, 1 provider each; webagent serve does the rest Every tool call passes through action.
Guard before execution Live today: OpenRouter or gateway brains, MCP tools, Slack, WhatsApp, HTTP channels Not yet: browser actions, OAuth MCP, OTel export, AgentNet billing Apache 2.0, Go, v0; read the deferred-hardening list before production Check out the GitHub repo and the launch post.
All credit goes to the researcher of this project.Also, feel free to follow us on Twitter and don’t forget to join our 150k+ML SubReddit and Subscribe to our Newsletter.Wait!are you on telegram?now you can join us on telegram as well.
Need to partner with us for promoting your GitHub Repo OR Hugging Face Page OR Product Release OR Webinar etc.?Connect with us The post Agent-net Open Sources Webagent: A Go Harness That Turns Any Website into a Guarded AI Agent appeared first on MarkTechPost.
Related
相關文章

美國 AI 巨頭提議開發減速,歐洲同行和政界並不認同
作者:清源 責編:清源 評論: 9 月 18 日消息,據路透社今天(18 日)報道,此前,阿莫迪、奧爾特曼和馬斯克先後發出警告,能力不斷增強的 AI 系統可能帶來風險,故有必要控制其發展節奏,但歐洲企業和官員對此表現出明顯懷疑。法國初創企業 Mistral 在聲明中指出,這些風險早在幾個月前就已十分明確了。

MiniMax Code CLI 正式開源,在評測中取得 76.7% 的任務通過率
作者:沁滄(實習) 責編:沁滄 評論: 感謝網友 Domado 的線索投遞!9 月 18 日消息,MiniMax 今晚宣佈,MiniMax Code CLI 的 v0.4.12 版本面向全球開發者正式開放,並且以 MIT 協議正式開放源代碼。

北京發佈“詞元經濟十條”:高標準建設詞元工廠,推動關鍵核心技術攻關
作者:浩渺 責編:浩渺 評論: 感謝網友 蛋殼兒 的線索投遞!9 月 18 日消息,北京市經濟和信息化局今日宣佈,北京市“詞元經濟十條”正式發佈。為貫徹落實《國務院關於深入實施“人工智能+”行動的意見》(國發〔2025〕11 號),率先培育智能經濟新形態,以詞元(Token)為抓手,發展詞元經濟新增量,制定《北京市加快詞元經濟發展的行動方案(2026—2028 年)》(注:以下簡稱《行動方案》)。
PrismML Releases Ternary Bonsai 2 27B: A 5.9 GB Apache 2.0 Model Retaining 98.2% of Qwen3.8 27B Performance
PrismML has released Ternary Bonsai 2 27B, a ternary-weight version of Qwen3.8 27B. The language model occupies 5.93 GB, against 53.80 GB in FP16. PrismML reports that it keeps 98.2% of the parent model’s average across 20 benchmarks.

OpenAI剛曝光循環架構,這家公司更早將其用於世界模型
鍵詞只有一個:循環。 Astra採用了一種被稱為“循環深度”(Recurrent Depth)的架構,本質是讓同一組Transformer層被反覆複用,用更少的參數實現更深的計算。 這被外界視為OpenAI對傳統“堆參數、堆算力”路線的一次重大修正——不再只是把模型做大,而是讓模型學會“反覆思考”。 消息一齣,整個AI社區迅速升溫。

千問辦公接入高德門店經營專家套件 提升實體店選址與經營效率
此項新功能自9月18日起正式上線,用戶只需在千問辦公添加相應套件並完成高德賬號授權,即可實現從門店選址到日常經營分析的全鏈路工作。隨著實體商業的不斷發展,傳統的人工選址和經營分析方式顯得效率低下且容易出錯。個體創業者在開新店時,往往需要花費大量時間進行實地勘察,整體耗時可達一週,而商家在監測門店熱度變化及競爭對手動態時,也面臨數據整理繁瑣、分析結果不準確的問題。