Skip to main content
ClaudeWave
EN, English
parcadei / Subagents

Continuous-Claude-v3

Context management for Claude Code. Hooks maintain state via ledgers and handoffs. MCP execution without context pollution. Agent orchestration with isolated context windows.

  • 3.9k stars
  • 302 forks
  • 47 open issues
  • Language: Python
  • License: MIT
  • updated
Install in ClaudeView on GitHub

Install

Terminal
git clone https://github.com/parcadei/Continuous-Claude-v3 && cp Continuous-Claude-v3/*.md ~/.claude/agents/
  1. Clone the repository and copy the agent .md definitions into ~/.claude/agents (or .claude/agents inside a project).
  2. Start a new Claude Code session to load the agents.
  3. Delegate work to them with the Task/Agent tool or by name.

Editor's note

Continuous Claude is a context management and agent orchestration framework built on top of Claude Code that addresses the platform's context compaction problem, where accumulated conversation history is collapsed and nuanced session decisions are lost. It does this through three interlocking systems: 109 natural-language-triggered skills, 32 specialized subagents (including scout for codebase exploration and oracle for research), and 30 hooks that fire automatically during the Claude Code lifecycle. A YAML-based handoff system transfers state between sessions more token-efficiently than prose summaries, while a background daemon extracts learnings after each session and surfaces them on resumption. The TLDR code analysis layer performs five-pass semantic indexing so Claude reads compressed file representations rather than raw source, reducing token burn. Hooks also run pyright and ruff immediately after edits to catch errors before tests run. Users interact in plain language rather than slash commands, and a hook injects skill-activation suggestions into each message. The system targets developers doing long, multi-session projects in Claude Code who need compounding institutional memory rather than perpetual cold starts.

Installable items

24 items in this repository

aegisSubagent

Security vulnerability analysis and testing

Install
agent-context-isolationSkill

Agent Context Isolation

Install
agent-orchestrationSkill

Agent Orchestration Rules

Install
agentic-workflowSkill

Agentic Workflow Pattern

Install
agentica-agentSubagent

Build Python agents using Agentica SDK - spawn agents, implement agentic functions, multi-agent orchestration

Install
agentica-claude-proxySkill

Guide for integrating Agentica SDK with Claude Code CLI proxy

Install
agentica-infrastructureSkill

Reference guide for Agentica multi-agent infrastructure APIs

Install
agentica-promptsSkill

Write reliable prompts for Agentica/REPL agents that avoid LLM instruction ambiguity

Install
agentica-sdkSkill

Build Python agents with Agentica SDK - @agentic decorator, spawn(), persistence, MCP integration

Install
agentica-serverSkill

Agentica server + Claude proxy setup - architecture, startup sequence, debugging

Install
agentica-spawnSkill

Spawn Agentica multi-agent patterns

Install
arbiterSubagent

Unit and integration test execution and validation

Install
architectSubagent

Feature planning, design documentation, AND integration planning

Install
ast-grep-findSkill

AST-based code search and refactoring via ast-grep MCP

Install
async-repl-protocolSkill

Async REPL Protocol

Install
atlasSubagent

End-to-end and acceptance test execution

Install
background-agent-pingsSkill

Background Agent Pings

Install
braintrust-analystSubagent

Analyze Claude Code sessions using Braintrust logs

Install
braintrust-analyzeSkill

Analyze Claude Code sessions via Braintrust

Install
braintrust-tracingSkill

Braintrust tracing for Claude Code - hook architecture, sub-agent correlation, debugging

Install
buildSkill

Workflow orchestrator that chains existing skills for feature development

Install
chroniclerSubagent

Session analysis, precedent lookup, and learning extraction

Install
cli-referenceSkill

Claude Code CLI commands, flags, headless mode, and automation patterns

Install
commitSkill

Create git commits with user approval and no Claude attribution

Install
README

Subagents overview

README preview not available. Visit the repo on GitHub for full documentation.
Topics
agentsclaude-codeclaude-code-cliclaude-code-hooksclaude-code-mcpclaude-code-skillsclaude-code-subagentsclaude-skillsmcp
Frequently asked

What people ask about Continuous-Claude-v3

What is parcadei/Continuous-Claude-v3?

parcadei/Continuous-Claude-v3 is subagents for the Claude AI ecosystem. Context management for Claude Code. Hooks maintain state via ledgers and handoffs. MCP execution without context pollution. Agent orchestration with isolated context windows. It has 3.9k GitHub stars and its last recorded update is dated 2026-01-26.

How do I install Continuous-Claude-v3?

You can install Continuous-Claude-v3 by cloning the repository (https://github.com/parcadei/Continuous-Claude-v3) or following the README instructions on GitHub. ClaudeWave also provides quick install blocks on this page.

Is parcadei/Continuous-Claude-v3 safe to use?

Our security agent has analyzed parcadei/Continuous-Claude-v3 and assigned a Trust Score of 77/100 (tier: Trusted). See the full breakdown of passed checks and flags on this page.

Who maintains parcadei/Continuous-Claude-v3?

parcadei/Continuous-Claude-v3 is maintained by parcadei. The last recorded GitHub activity is dated 2026-01-26, with 47 open issues.

Are there alternatives to Continuous-Claude-v3?

Yes. On ClaudeWave you can browse similar subagents at /categories/agents, sorted by popularity or recent activity.