#256 · Primary category: AI Coding Assistants
zerostack
Lightweight coding agent written in Rust, optimized for memory footprint and performance
Project last updated:08/27/26
GitHub Stars
1.6K
Forks
125
Contributors
29
License
GPL-3.0
Why we included this project
zerostack is a terminal-based coding agent in Rust that keeps memory use low without dropping tools. It works with OpenAI, Anthropic, Gemini, and local Ollama models, exposes the standard coding-agent tool set, and can add more through MCP servers, so it fits into existing workflows instead of demanding a new one. Developers who have felt heavier JavaScript agents strain their machines will appreciate the roughly 16 MB footprint and sub-second startup, which make it reasonable to run several agents at once. The operational side is covered too, with permission modes and per-tool rules, runtime-switchable prompts for planning, review, or debugging, session save and resume with auto-compaction, git worktrees for branch-per-task work, and optional persistent Markdown memory. And if you work in an editor like Zed, the optional Agent Communication Protocol support lets it drive zerostack directly as the coding-agent backend.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
superpowers
An agentic skills framework & software development methodology that works.
ECC
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
skills
Skills for Real Engineers. Straight from my .agents directory.
opencode
The open source coding agent.
andrej-karpathy-skills
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.