Docker

Headlines · 5

1. 2d
   Open-source multi-agent pentest system PentAGI hits GitHub Trending

   PentAGI, listed on GitHub Trending, is an open-source, self-hosted multi-agent system for autonomous penetration testing from VXControl, a security engineering company based in Dubai: a user describes a target and a goal in plain language, and specialized AI agents plan and carry out the work, running 20-plus professional tools such as nmap, Metasploit and sqlmap inside an isolated Docker sandbox before producing a report. The project says it supports 10-plus cloud LLM providers (OpenAI, Anthropic, Gemini, AWS Bedrock, DeepSeek, Qwen and others) plus local models via Ollama or vLLM, and ships with PostgreSQL and pgvector long-term memory, an optional Neo4j/Graphiti knowledge graph, execution monitoring and task planning, REST and GraphQL APIs, and Grafana/Langfuse observability; an enterprise edition adds agent profiles, plugins, MCP servers and a real-time attack-chain view. The maintainers state that PentAGI today is an autonomous or assistant-guided pentest platform, not a CALDERA-style breach-and-attack-simulation product with predefined campaigns, and that it is for authorized testing only.

   GitHub Trending（每日） · 🔥 5
2. 2d
   Intel launches inference microservices for Xeon processors

   Intel announced Intel Inference Microservices, a containerized microservice offering for large language model inference aimed at letting enterprises get tuned inference on Xeon processors they already own without days of manual hardware optimization. Each microservice is a standalone container serving one model over an OpenAI-compatible API; at startup it detects the processor generation, assigned cores, NUMA layout and built-in accelerators such as AMX, AVX-512, AVX2 and VNNI, then applies a validated configuration profile. The first versions are based on the vLLM engine with BF16 precision, run in Docker or Kubernetes 1.24+ on 4th, 5th or 6th Gen Intel Xeon with 32GB RAM suggested. Intel says later versions will add Helm charts, a Kubernetes Operator, more engines and precisions, and that the same model will extend to Intel GPUs including the Crescent Island data-center inference GPU.

   IT之家 AI · 🔥 0
3. 4d
   Armature study: how Claude Code, Codex and Cursor pick tools

   Armature, a developer-tool growth services company founded in 2026, published a study of how coding agents choose third-party tools, testing Claude Code, Codex and Cursor across 75 public repositories with 1,163 prompts built to simulate different developer personas; the company says its public core figures come from a verified subset of 5,292 sessions out of roughly 17,000 in total. Repository context mattered heavily: for email/communications providers, Resend won in TypeScript (55 of 89 runs), SendGrid in Python (22 of 24), Postmark in Go (20 of 24) and Azure ACS in Java (22 of 23), while the three agents picked the same tool in only 42% of test units and Claude Code chose to build something itself almost twice as often as Codex and Cursor (19% vs 10%). The agents also drew on different information sources — Codex used web search in 94% of sessions, Cursor relied on web information in about two-thirds, and Claude Code leaned mainly on its own knowledge, searching only around 30% of the time but visiting three times as many pages as Codex when it did browse.

   InfoQ 中文 AI&大模型 · 🔥 0
4. 5d
   Carbonato botnet hijacks unsecured Docker hosts, deploys AI agent to steal credentials

   Singapore's Cyber Security Agency (CSA) issued an advisory warning that the Carbonato botnet is compromising hosts through unprotected Docker remote management interfaces. According to the advisory and research by security firm ThreatDown, the campaign does not exploit a new Docker flaw but instead targets the Docker Remote API exposed over TCP port 2375 without authentication; attackers create privileged containers to take control of the underlying host and maintain remote access via mechanisms such as SSH reverse tunnels. ThreatDown says the intruders then deploy the open-source AI agent framework Hermes Agent, sending tasks remotely over Telegram so a large language model can interpret them, generate commands and run them on the victim host, with the agent specifically instructed to prioritize collecting API keys for AI services and to look for SSH keys, access tokens and database information. CSA recommends keeping Docker remote interfaces off the public internet, using SSH or client-authenticated TLS instead, and isolating affected hosts and rotating potentially exposed credentials before restoring service.

   iThome 台湾 · 🔥 0
5. Oct 8
   Docker ships Docker Agent: YAML-defined agents with MCP support

   Docker has released Docker Agent, an official tool for building and running AI agents from declarative YAML configs, with built-in tools and multi-agent orchestration. It ships as a docker CLI plugin, usable via the docker agent command after installing Docker Desktop 4.63 or later, and is also installable through Homebrew. The tool is said to support arbitrary MCP servers and to distribute and share agents as OCI images.

   开源中国 · 🔥 0

Experience and discussion from the community

Share my Docker experienceAsk about Docker