🤖 augustus

LLM security testing framework for detecting prompt injection, jailbreaks, and adversarial attacks. See also the companion blog post.

https://github.com/praetorian-inc/augustus

#AI

🤖 3 Principles for Designing Agent Skills

Block Engineering discusses designing agent skills using three principles: make deterministic outputs script-based, let agents handle interpretation and conversation, and write explicit constitutional constraints. Skills codify tribal knowledge into executable documentation for AI agents across their organization.

https://engineering.block.xyz/blog/3-principles-for-designing-agent-skills

#AI

🔶🤖 Building an AI-powered defense-in-depth security architecture for serverless microservices

This AWS blog demonstrates implementing a seven-layer AI-powered defense-in-depth security architecture for serverless microservices using AWS Shield, WAF, Cognito, API Gateway, VPC, Lambda, Secrets Manager, and DynamoDB, enhanced with GuardDuty and Amazon Bedrock for intelligent threat detection and automated response.

https://aws.amazon.com/ru/blogs/security/building-an-ai-powered-defense-in-depth-security-architecture-for-serverless-microservices/

(Use VPN to open from Russia)

#aws #AI

🤖 MCP Server Security: The Hidden AI Attack Surface

MCP servers connecting AI assistants to external tools create significant attack surfaces enabling arbitrary code execution, data exfiltration, and social engineering. Both local and remote MCP servers can be exploited through server chaining, supply chain attacks, and malicious tool implementations.

https://www.praetorian.com/blog/mcp-server-security-the-hidden-ai-attack-surface/

#AI

🤖 Using threat modeling and prompt injection to audit Comet

Trail of Bits used ML-centered threat modeling and adversarial testing to identify four prompt injection techniques that could exploit Perplexity's Comet browser AI assistant to exfiltrate private Gmail data. The audit demonstrated how fake security mechanisms, system instructions, and user requests could manipulate the AI agent into accessing and transmitting sensitive user information.

https://blog.trailofbits.com/2026/02/20/using-threat-modeling-and-prompt-injection-to-audit-comet/

#AI

🤖 How "Clinejection" Turned an AI Bot into a Supply Chain Attack

A prompt injection in a GitHub issue title gave attackers code execution inside Cline's CI/CD pipeline, leading to cache poisoning, stolen npm credentials, and an unauthorized package publish affecting the popular AI coding tool's 5 million users. Here's the full technical breakdown and what developers should do now.

https://snyk.io/blog/cline-supply-chain-attack-prompt-injection-github-actions/

(Use VPN to open from Russia)

#AI

🤖 Running OpenClaw safely: identity, isolation, and runtime risk

OpenClaw, a self-hosted agent runtime, lacks built-in security controls, enabling credential exfiltration, memory/state manipulation, and host compromise via indirect prompt injection and malicious skills. Microsoft recommends isolated deployment, least-privilege identities, continuous monitoring, and Defender XDR hunting queries.

https://www.microsoft.com/en-us/security/blog/2026/02/19/running-openclaw-safely-identity-isolation-runtime-risk/

#AI

🤖 hackerbot-claw: An AI-Powered Bot Actively Exploiting GitHub Actions

A week-long automated attack campaign targeted CI/CD pipelines across major open source repositories, achieving remote code execution in at least 4 out of 5 targets. The attacker, an autonomous bot called hackerbot-claw, used 5 different exploitation techniques and successfully exfiltrated a GitHub token with write permissions from one of the most popular repositories on GitHub. This post breaks down each attack, shows the evidence, and explains what you can do to protect your workflows.

https://www.stepsecurity.io/blog/hackerbot-claw-github-actions-exploitation#attack-6-aquasecuritytrivy---evidence-cleared

#AI

🤖 The Reach Pattern

The "Reach" pattern is a personal CLI that hijacks existing browser sessions to query SaaS APIs (Slack, Jira, Confluence, etc.) on your behalf, feeding structured organizational context to your AI coding assistant.

https://jackdanger.com/the-reach-pattern

#AI