Composable Security ⛓️💥 published an article detailing practical isolation options for AI coding agents, including hardened local setups, Dev Containers, VMs, and ephemeral environments.
Public source
Publisher name
Public post
We’ve all heard stories about AI agents escaping sandboxes or finding unexpected ways around supposedly isolated environments. Recent incidents, including the OpenAI / H…
Company
Composable Security ⛓️💥
Smart Contracts, Off-chain Components and AI Integrations Security Reviews Clients: Lido,UniswapFND,RedStone,YieldNest
- Industry
- Blockchain Services
- Location
- KRAKÓW, PL
- Company size
- 2–10 employees
About Composable Security ⛓️💥
We've already helped dozens of projects keep their products and systems secure, but we believe we can do even more. Participation in conferences such as ETHcc or ETHWarsaw and numerous interactions with the community convinced us that blockchain-based projects are the future. It is not just about technology, it’s about the people who create it. Seeing how committed, open-minded and willing to educate web3 developers are, we decided to go full time and do everything we can to help them build a better, more efficient and more transparent world. The Composable Security company was founded in August 2022 by Damian Rusinek (@drdr_zz) and Paweł Kuryłowicz (@wh01s7) who have actively contributed to the crypto space since 2017. Starting from the creation of the Smart Contract Security Verification Standard, through numerous audits of smart contracts, to the introduction of threat modeling to DeFi. Through numerous publications, presentations, and a series of educational tweets, we tried to share the best security practices developed thanks to many years of experience in web2. Security should be adapted to each stage of development, security should be composable to the best suits visionary needs. That’s why we started Composable Security. Currently provided services: - Smart contract security reviews/audits - Threat modeling - Threat analysis - Security consultation If you are interested, contact us through info@composable-security.com
See moreLatest activity
Latest activity from Composable Security ⛓️💥
3 signals
Discover more
Similar signals
Similar public activity from other companies.
Research & Knowledge
Tatum
Tatum published a post breaking down the risks of connecting AI agents to on-chain tools, the economic validity of authorization, and the need for layered defenses using isolated KMS, real-time webhooks, and independent state validation.
Research & Knowledge
Lineaje
Lineaje published a research paper titled AI TAR VIDEO: An AI coding agent turned a routine repository read into a secret exfiltration pipeline — no jailbreak required, just the privilege it already had.
Research & Knowledge
Joyful Agents
Joyful Agents published a paper titled Temporal Dynamics of Memory Poisoning in Web3-Style LLM Agents that constructs a schema-constrained dataset of 2,614 multi-step attack trajectories spanning four attack families, chain poisoning, policy rewriting, backdoor triggering, and slow drift executed over shared persistent memory.
Research & Knowledge
Grip Security
Grip Security published an article looking beyond prompt injection at the full agentic AI attack surface and controls security teams need to limit what happens when an agent gets it wrong.
Research & Knowledge
Chainloop