7-e1even
Featured
learn-agent
A collection of notes on building coding agents, derived from the production development of the desktop agent Reina, with each mechanism simplified into a zero-dependency, single-file Node.js demo.
poirot
Poirot is a deep research agent kernel with a middleware-first architecture, five-layer long-term memory, multi-agent orchestration, sandbox isolation, and a three-layer skill self-evolution system, backed by 2400+ tests.
Stars
10
7-day growth
No data
Forks
0
Open issues
0
License
MIT
Last updated
2026-07-28
Why it is worth attention
It treats agent internals—memory, skill injection, tool routing, and context governance—as pluggable middleware, not embedded logic, enabling independent testing and modular evolution.
Who it is for
- Developers building custom research agents or multi-agent systems
- AI researchers interested in cognitive‑science‑inspired memory architectures
- Engineers needing sandboxed, multi‑LLM fallback, and observable agent pipelines
- System architects evaluating modular agent frameworks for production
Use cases
- Conducting long‑running, multi‑step research with persistent memory across sessions
- Delegating sub‑tasks (e.g., coding) to specialized agents sharing a sandbox
- Building skill sets that self‑evolve via LLM‑driven mutation and quality gating
- Enforcing sandboxed file isolation and path whitelisting for agent tool execution
Strengths
- Decoupled middleware design (21 middleware) cross-cutting every lifecycle hook
- Five‑layer memory with lazy Ebbinghaus decay, pure BM25 retrieval, and non‑blocking auto‑consolidation
- Real context‑window resolution through `FallbackChatModel` for accurate token budgeting
- Over 2400 tests guarding every module, ensuring independent verifiability
Considerations
- Requires Python 3.12+ and a configured DeepSeek API key as the ultimate fallback
- Production sandbox isolation depends on Docker, adding infrastructure overhead
- Configuration surface is large (environment variables, MCP servers, skill settings) which may overwhelm simple use cases
README quick start
Quick Start
Description
Poirot is a deep research agent kernel built for those who care about how agents are architected.
Related repositories
Similar projects matched by category, topics, and programming language.
deerwork-ai
Featured
deer-workflow
An open-source Dynamic Workflow runtime that combines deterministic TypeScript orchestration with replaceable Agent runtimes.
makecindy
Featured
cindy
Cindy is an open-source AI agent that runs locally on your machine, integrates multiple AI harnesses and models, and provides memory, skills, and automation to perform real work in your projects and apps.