Introducing Secure ClawdBot with access to 1000+ apps! We've merged Claude Code and Composio to create a ClawdBot that: - securely uses 1000+ apps with zero setup - is easy to set up and use via terminal - proactively interacts with you on WhatsApp, Telegram, Signal Free and open source:

Genuine question, but what have you built here? The readme is verbose and AI-generated, not seeing anything about how this is secure and what it means to be integrated with ‘composio’

What exactly makes it a "secure" ClawdBot?

What secure means: • Remote sandboxed execution with tool router – any code runs in an isolated remote environment, not on your local machine. • All third-party access goes through Composio – credentials are never handled directly by the bot or Claude; integrations are mediated centrally. • Managed scopes – you explicitly control what each app can do (read, write, limited actions), and can revoke access anytime.

The 'secure' part is what everyone's asking about in the replies, connecting 1000+ apps sounds powerful but also kinda terrifying lol. What's the actual security model here? OAuth tokens stored locally or something else?

As long as it doesnt work with our 200/90/20 € Claude coding plans, it will be hard to accept these expensive API based apps. Molt runs on MinimaxM2 at 10€/month and can run CC cli ...

nice work!

Securely using 1000+ apps with zero setup sounds impressive, but doesn't that also mean you're introducing a single point of failure if the ClawdBot is compromised. How do you plan to mitigate that risk, especially when interacting with sensitive apps.

PSA: if you are going to compete with clawdbot, implement a 1 click migration.

builders who make their ai agents work with existing tools instead of reinventing everything are gonna move way faster. this is exactly how it should work

Composio is great but not a huge fan of data retention policy though Not sure this means

@grok how can i use it, explain step by step

cannot run with ollama? oh, no, thanks.

how is this secure, would oyu link it to your presonal email...?

Was thinking when Composio would drop this

awesome to see

👏

🔥🦞🦀

this could actually replace me

Rube?

composio integration is a smart move. the biggest pain with clawdbot has been wiring up individual app connections manually. having oauth handled out of the box for 1000+ services removes the main friction point. the security questions in the replies are valid though, would love to see a deep dive on how tokens are scoped and stored locally vs cloud

@calebrebelo_ note

Did you mean claude cowork?

1000+ apps with that 'secure' part? Have you thoroughly tested those? There are specific issues that can't be solved like architectural limits, where injection attacks can happen through a friendly email, essentially tricking whatever LLMs that runs those.

How do you handle running out of API credits?

@grok it is safe?

Secure ClawdBot + Composio is a massive level-up. Proactive interaction on Telegram/Signal is exactly where the local agent UX needs to go. Great work on this merger! 🤖🔥

@bankrbot deploy Secure ClawdBot with ticker SeClawd

The 1000+ app access is the real unlock. Agents need tool breadth, not just depth. Composio integration is smart — modular connectors beat monolithic plugins every time. Question: how does auth delegation work across apps? That's usually where security gets messy.

Pretty awesome!

The headline contradicts itself. Enjoying clickbait. They clearly have no idea what they’re saying or doing

Are you following @asharoraa

Amazing!!🔥

incredible

having an agent that actually pings me on signal when a task is done is exactly what the dev ux was missing tbh

It’s no longer called ClawdBot. Expect 75 pages of legal documents from Anthropic