backbay-labs

backbay-labs / clawdstrike

Public

Runtime security enforcement and threat hunting engine for autonomous agent fleets. Kernel to chain. Proof, not logs. Build Swarm Detection & Response (SDR) platforms with Clawdstrike.

backbay.io agent-security agentic-security agents cyber-defense cyber-security
185
17
100% credibility
Found Feb 04, 2026 at 43 stars 4x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

Clawdstrike is a security tool that enforces policies on AI coding agents to prevent risky actions like accessing secrets or suspicious network calls, with a desktop app for monitoring and a marketplace for policies.

How It Works

1
🔍 Discover Clawdstrike

You hear about Clawdstrike, a friendly security helper that watches your AI coding tools to keep them safe from risky moves.

2
📥 Install the agent

Download and run the simple agent app, which quietly sets up protection in the background with a tray icon for easy control.

3
🔗 Link your AI tool

Click to connect your favorite AI coder like Claude or Cursor, so it asks permission before touching files or the internet.

4
🖥️ Open the dashboard

Launch the colorful desktop view to see live events, test rules, and explore shared safety setups from others.

5
👀 Watch it work

As you code with AI, see real-time alerts for blocked dangers, with proof receipts showing exactly what happened.

✅ Stay secure effortlessly

Your AI tools now run safely, you monitor everything easily, and you feel confident creating without worries.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 43 to 185 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is clawdstrike?

Clawdstrike lets you build your own Swarm Detection & Response (SDR) platform and OpenClaw security infrastructure in Rust, enforcing runtime policies on AI agents to block risky actions like file access, network egress, or jailbreaks. It generates signed receipts for every decision, ensuring tamper-proof audits, and integrates via SDKs for TypeScript, Python, and WebAssembly. Developers get a fail-closed guard system with built-in detectors for secrets, patches, and prompt injection, plus apps for desktop monitoring and agent tray enforcement.

Why is it gaining traction?

Its sub-50µs overhead per check makes it invisible during LLM tool calls, unlike heavier EDR tools, while multi-framework hooks for Claude Code, Cursor, and LangChain simplify securing AI coding agents. Jailbreak detection layers (heuristics to ML) and prompt watermarking stand out for devs building own AI or GitHub Copilot agents. The open policy YAML schema and CLI for quick tests hook users fast.

Who should use this?

Security engineers hardening AI agents in production, like those building own LLM pipelines or MCP servers for Cursor/Cline. Devs creating GitHub Actions or apps with tool boundaries, or teams assembling SDR fleets on OpenClaw. Ideal if you're building own NAS/router/PC with agent ops and need signed proofs without perf hits.

Verdict

Alpha software with solid docs, benchmarks, and Tauri apps, but only 74 stars and 1.0% credibility signal early-stage risks—test in sandboxes first. Grab it if you're prototyping AI security today; production teams wait for stable releases.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.