Havuz

WAKE.md for AI agents: compile project state so agents stop starting cold.

Exploring agent "thinking" animation.

See where your Claude Code tokens go — cost per call scales 4.3x with context size. Hooks + SQLite + dashboard + MCP server for context wind

CTXORA Engine — Local-first context engine for coding agents. Index once. Ground every agent.

给你的AI订一份晨报：常驻agent的信息摄入与跨context认知循环（思路篇，无代码）

Gates 84.7% of tool-schema prompt tokens away for DeepSeek Harness: plugins declare themselves organs, a nerve impulse routes each command,

Staged context pruning for OpenCode, Oh My Pi, pi, and Claude Code.

Local task-state ledger for AI coding agents. Preserve decisions, source conclusions, evidence, and checkpoints across Codex, Claude Code, C

Git-native decision memory for Claude Code, Codex, Cursor and other coding agents. Preserve constraints and rejected alternatives; surface o

Token-efficient Windows file discovery for Codex and AI agents using voidtools Everything (es.exe)

Entente — provenance engineering for coding agents: handoffs become confirmable, verifiable, retryable contracts; agents form a network with

CLI + MCP for agent context: a task layer over TickTick, Notion, GitHub, Obsidian & more (RRF retrieval, provenance, reviewed writes, undo)

Token-budgeted codebase discovery for agent shells and Pi: cross-language maps, focused navigation, and impact analysis.

Wordcell keeps decisions, plans, and sources as Markdown beside your code, so coding agents can find them from the file they are about to ch

The memory layer that gets better as you use it. What broke, why, what depended on what, even the "push this to Wednesday" kind of thing tha

Agent memory that needs no API key, no LLM, and no embedding provider. One command on Bun and it is serving MCP over stdio against a local S

Context layer for AI agents in vertical SaaS: domain events in, a bitemporal context graph per end-customer out, five MCP tools for every ag

Claude Code memory system — auto context injection, drift detection, compounding, FTS5/vector search, Obsidian vault export. Zero core deps.

Lossless context folding for the Pi coding agent: only the oldest context changes; the fresh window stays untouched.

Logic programming for AI agents: Datalog-family language compiling to optimized SQL

⚡ TidyFactor Architecture Ecosystem , AI-Native Web Architecture, Design Systems & Agent Skills .

A small, file-first verification harness that turns an AI agent's "I'm done" claim into an auditable admission decision. 100% offline & zero

Offline-first toolkit for agent-ready repositories: readiness scoring, instruction graphs, context packs, policy/rule packs, MCP, GitHub Act

AI remembers, you find. File-based long-term memory and project maps for any AI coding tool, plus a zero-dependency governance CLI for chang

🐀 Terminal-native AI data agent — ask your warehouse a question in plain English, watch dialect-correct SQL stream into your terminal, get g

A structured agent runtime for LLMs. Structured context, multi-stage workflows, and ECS-based orchestration in a single binary

Gobstopper is an intelligent compaction proxy for Claude Code and Codex. It compacts more often, keeps history local, and cut tokens 29% on

Minimal harness-agnostic recursive language model layer — one binary, Python + llm()

Turn long AI conversations into reusable domain knowledge across sessions

Persistent coding intelligence for Codex — project-aware context, adaptive reasoning, scoped changes, and risk-aware verification.

Stop re-explaining your work to your coding agent. Plain markdown and YAML in git, read at session start by Claude Code, Codex and Copilot C

Knowledge engine for AI agents: your documents become a git-versioned markdown knowledge graph the agent navigates over MCP. Curated summari

A warn-only runtime layer for Claude Code agents. Keeps feature state, scope and handoffs on disk so they survive compaction, and requires a

Local-first repository context for coding agents: codegraph, cited project knowledge, change impact, and evidence-backed workflows.

The context exoskeleton for AI development: optimize the entire development lifecycle to reduce token usage and context occupancy.

面向 Vibe Coding 的能力增强型协作框架（Harness），让 Vibe Coding 看得见，可审查、可学习、可进化。