Athena-Public：跨 IDE 的可移植 AI 智能体上下文记忆工具

An agentic tool to compound your context over time, portable across IDEs.

Your context lives in plain Markdown on your disk, works across Claude Code, Antigravity, Cursor, Gemini CLI and VS Code — persistent memory, structured reasoning, and governed AI agents that work across any LLM.

Quickstart · How It Works · Docs · Why Athena? · Safety

The Problem

You've spent months training ChatGPT to understand you. Then a model update resets the personality. You switch to Claude or Gemini — you start from zero.

Platform memory is unreliable, opaque, and locked to one provider. You don't own it and you can't take it with you.

Athena moves the memory layer to your machine: plain Markdown files that you own, version-control, and point at any model. The model is just whoever's on shift.

Quickstart

git clone https://github.com/winstonkoh87/Athena-Public.git && cd Athena-Public
python3 -m venv .venv && source .venv/bin/activate
pip install -e ".[local]"               # lightweight — no cloud deps
athena init --ide claude                # or: antigravity, cursor, gemini, vscode, kilocode, roocode
athena doctor                           # expect 0 failures

Then type /start in your IDE's AI chat panel. Work normally. Type /end to save.

Full install (cloud sync + reranking): pip install -e ".[full]" See Getting Started for Windows, advanced config, and Supabase setup.

How It Works

┌─────────────────────────────────────────────────────┐
│  Your IDE (Claude Code / Antigravity / Cursor / …)  │
└───────────────────┬─────────────────────────────────┘
                    │ hooks (code-enforced, not prompt-based)
                    ▼
┌─────────────────────────────────────────────────────┐
│  Athena SDK                                         │
│  ├── Lifecycle: /start loads ~2K tokens, /end saves │
│  ├── Memory: session logs + canonical facts         │
│  ├── Search: hybrid (keyword + optional vectors)    │
│  └── Guardrails: ruin check, secret scan, grounding │
└───────────────────┬─────────────────────────────────┘
                    │
                    ▼
┌─────────────────────────────────────────────────────┐
│  Your Disk (plain Markdown — git-versioned)         │
│  └── Optional: Supabase pgvector for cloud sync     │
└─────────────────────────────────────────────────────┘

Three modes, one loop:

What's enforced in code vs. by prompt

Most AI-agent READMEs state every claim in the same confident voice. This one doesn't.

Why publish this table? Because the failure mode of this product category is self-mythologizing — describing aspirations in the present tense. Athena's own convention (Epistemic Status) requires labeling every mechanism as code-enforced, agent-discretion, or aspirational. This table is that convention applied to the README.

Measured, Not Claimed

# Run the tests yourself
pytest tests/ -v --tb=short

# Run the local KV cache & TTFT benchmark (Ollama / llama.cpp)
python scripts/benchmark_kv_cache.py --model qwen2.5-coder:14b

# Run the retrieval evaluator yourself (requires Supabase keys)
python examples/scripts/evaluator.py --gold-set .agent/eval/gold_set.json

Anti-Goodhart Invariant: Why did our reported Hit@5 shift from 0.89 to 0.57? Lenient substring matchers count partial word overlaps as "hits," inflating benchmark scores by ~36% without improving retrieval. We killed the lenient matcher because vanity metrics mask regressions. See the full breakdown: Anti-Goodhart Benchmarking in RAG.

Agent Compatibility

Documentation

Contributing

PRs welcome. Please read CONTRIBUTING.md first.

The codebase uses ruff for linting, pytest for tests, and CI must pass before merge.