Continuous-Claude-v3
starsforks
How to install Continuous-Claude-v3?
parcadei/continuous-claude-v3
$git clone https://github.com/parcadei/Continuous-Claude-v3> A persistent, learning, multi-agent development environment built on Claude Code
README.md
| 1 | # Continuous Claude |
| 2 | |
| 3 | > A persistent, learning, multi-agent development environment built on Claude Code |
| 4 | |
| 5 | [](LICENSE) |
| 6 | [](https://claude.ai/code) |
| 7 | [](#skills-system) |
| 8 | [](#agents-system) |
| 9 | [](#hooks-system) |
| 10 | |
| 11 | **Continuous Claude** transforms Claude Code into a continuously learning system that maintains context across sessions, orchestrates specialized agents, and eliminates wasting tokens through intelligent code analysis. |
| 12 | |
| 13 | ## Table of Contents |
| 14 | |
| 15 | - [Why Continuous Claude?](#why-continuous-claude) |
| 16 | - [Design Principles](#design-principles) |
| 17 | - [How to Talk to Claude](#how-to-talk-to-claude) |
| 18 | - [Quick Start](#quick-start) |
| 19 | - [Architecture](#architecture) |
| 20 | - [Core Systems](#core-systems) |
| 21 | - [Skills (109)](#skills-system) |
| 22 | - [Agents (32)](#agents-system) |
| 23 | - [Hooks (30)](#hooks-system) |
| 24 | - [TLDR Code Analysis](#tldr-code-analysis) |
| 25 | - [Memory System](#memory-system) |
| 26 | - [Continuity System](#continuity-system) |
| 27 | - [Math System](#math-system) |
| 28 | - [Workflows](#workflows) |
| 29 | - [Installation](#installation) |
| 30 | - [Updating](#updating) |
| 31 | - [Configuration](#configuration) |
| 32 | - [Contributing](#contributing) |
| 33 | - [License](#license) |
| 34 | |
| 35 | --- |
| 36 | |
| 37 | ## Why Continuous Claude? |
| 38 | |
| 39 | Claude Code has a **compaction problem**: when context fills up, the system compacts your conversation, losing nuanced understanding and decisions made during the session. |
| 40 | |
| 41 | **Continuous Claude solves this with:** |
| 42 | |
| 43 | | Problem | Solution | |
| 44 | |---------|----------| |
| 45 | | Context loss on compaction | YAML handoffs - more token-efficient transfer | |
| 46 | | Starting fresh each session | Memory system recalls + daemon auto-extracts learnings | |
| 47 | | Reading entire files burns tokens | 5-layer code analysis + semantic index | |
| 48 | | Complex tasks need coordination | Meta-skills orchestrate agent workflows | |
| 49 | | Repeating workflows manually | 109 skills with natural language triggers | |
| 50 | |
| 51 | **The mantra: Compound, don't compact.** Extract learnings automatically, then start fresh with full context. |
| 52 | |
| 53 | ### Why "Continuous"? Why "Compounding"? |
| 54 | |
| 55 | The name is a pun. **Continuous** because Claude maintains state across sessions. **Compounding** because each session makes the system smarter—learnings accumulate like compound interest. |
| 56 | |
| 57 | --- |
| 58 | |
| 59 | ## Design Principles |
| 60 | |
| 61 | An agent is five things: **Prompt + Tools + Context + Memory + Model**. |
| 62 | |
| 63 | | Component | What We Optimize | |
| 64 | |-----------|------------------| |
| 65 | | **Prompt** | Skills inject relevant context; hooks add system reminders | |
| 66 | | **Tools** | TLDR reduces tokens; agents parallelize work | |
| 67 | | **Context** | Not just *what* Claude knows, but *how* it's provided | |
| 68 | | **Memory** | Daemon extracts learnings; recall surfaces them | |
| 69 | | **Model** | Becomes swappable when the other four are solid | |
| 70 | |
| 71 | ### Anti-Complexity |
| 72 | |
| 73 | We resist plugin sprawl. Every MCP, subscription, and tool you add promises improvement but risks breaking context, tools, or prompts through clashes. |
| 74 | |
| 75 | **Our approach:** |
| 76 | - **Time, not money** — No required paid services. Perplexity and NIA are optional, high-value-per-token. |
| 77 | - **Learn, don't accumulate** — A system that learns handles edge cases better than one that collects plugins. |
| 78 | - **Shift-left validation** — Hooks run pyright/ruff after edits, catching errors before tests. |
| 79 | |
| 80 | The failure modes of complex systems are structurally invisible until they happen. A learning, context-efficient system doesn't prevent all failures—but it recovers and improves. |
| 81 | |
| 82 | --- |
| 83 | |
| 84 | ## How to Talk to Claude |
| 85 | |
| 86 | **You don't need to memorize slash commands.** Just describe what you want naturally. |
| 87 | |
| 88 | ### The Skill Activation System |
| 89 | |
| 90 | When you send a message, a hook injects context that tells **Claude** which skills and agents are relevant. Claude infers from a rule-based system and decides which tools to use. |
| 91 | |
| 92 | ``` |
| 93 | > "Fix the login b |
People who liked this also viewed
6 picksType
- fastmcp🚀 The fast, Pythonic way to build MCP servers and clients.MCP ServersJul 2026
- playwright-mcpPlaywright MCP serverMCP ServersJul 2026
- ext-appsOfficial repo for spec & SDK of MCP Apps protocol - standard for UIs embedded AI chatbots, served by MCP serversMCP ServersJul 2026
- langchain-mcp-adaptersLangChain 🔌 MCPMCP ServersJul 2026
- mcporterCall MCPs via TypeScript, masquerading as simple TypeScript API. Or package them as cli.MCP ServersJul 2026
- mcp-proxyA bridge between Streamable HTTP and stdio MCP transportsMCP ServersJul 2026
