Head of Ecommerce & Digital Marketing on Kenay Home & Banak & Klast 🚀 Anteriormente en @leroymerlin_es @endesa y @carrefourES 🏖 Con mi side-project @thrdemail
Give your AI agents a real voice. 🤖📧
Introducing https://t.co/V1iAwjKJNd: 100% autonomous email for agents. No SMS, no CAPTCHAs, no human friction.
Tier 1 Beta is LIVE! Just tell your agent:
"Go to https://t.co/D8pq7wRzBr and create your own email account."
#buildinpublic
@carlosvillu@thrdemail Si. Las IAs al contrario que nosotros, conocen perfectamente todos los códigos HTTP, por lo que cuando detectan un 402, sin son agentes autónomos van a buscar en la API como pueden salir de ahí. (Ahí ya depende lo fácil que se lo pongas)
@0interestrates I completely agree. We’ve to work on making the internet more machine-friendly. We’re building @thrdemail so that agents can be independent when sending and receiving emails (OTPs, logins, etc.). Hopefully, one day, agents will end up registering with Thrd and paying with Sponge.
cada inbound email se puntúa por señales como:
• intentos de override de rol
• petición de exfiltrar secretos
• “haz clic aquí / ejecuta esto”
• payloads ofuscados
Output determinista: allow / require ACK / block
qué canal te preocupa más, email o Slack/Teams?
Construyendo @thrdemail veo que el mayor riesgo no es el spam… es el prompt injection.
Gmail decide “spam / no spam”.
Un agente necesita algo más: “leer está OK, pero actuar puede ser peligroso”.
En Thrd estamos metiendo un Prompt Shield (sigue) #buildinpublic
@GSkrovina@Flynnjamm I am currently building an email infrastructure for AI agents. Zero configuration. Just one request to the agent and it already knows everything it needs to do 👉 https://t.co/V1iAwjKJNd
@AjaredIA This is the real enterprise wedge: not “chatbots”, but systems of record that surface the next best action inside the workflow. Insight → action → audit trail. That’s where adoption sticks.
@loganthorneloe 100%. “Agent performance” is mostly software engineering: retries, idempotency, state, evals, and observability. Once you treat agents like distributed systems, the wins compound fast.
Shipping in public: I just added real-time public stats — (and counting) + live inboxes created. It keeps me honest and it’s a forcing function for growth. What’s one metric you’d feel confident tracking publicly? #buildinpublic
@openclaw es una herramienta realmente impresionante. Sin embargo es insegura por lo que hay detrás (LLMs) por he creado Thrd. Una infraestructura de emails dedicada a Agentes de IA con medidas de seguridad por detrás como el Shield Prompt. Te dejo el link 👇
@Wenlopezn Precisamente por la misma paranoia de seguridad, es por la que estoy construyendo https://t.co/V1iAwjKJNd
Una infraestructura de email dedicada puramente a máquinas (Agentes) sí que tengamos que exponer nuestros emails y con varias capas de seguridad
Stop using your personal Gmail for AI agents. It’s a security risk. 🚩
Thrd. is the dedicated runtime for agentic email.
✅ Instant API provisioning
✅ Policy-gated outbound
✅ Total security isolation
Give your agents the inbox they deserve. 🤖📩
AI agents just got an upgrade. 🚀
The new 'thrd' skill on ClawHub allows agents to manage email accounts independently. Scale your agentic workflows with ease!
Download: https://t.co/v6BhJWWe9c
#OpenClaw#ClawHub#AI#MarketingTech
Humans ➡️ Gmail
Agents ➡️ thrd.
Finally, an email infrastructure built strictly for machines.
Get your API key: https://t.co/JoZRaM6dlo
#AIAgents#BuildInPublic