</>MCP Agents Market
Plugin

Continuous-Claude-v3

by parcadei3.9kPythonUpdated 2026-01-26

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

Claude Code

Continuous-Claude-v3 is a comprehensive Claude Code plugin that transforms Claude into a persistent, learning development environment with cross-session context management. It solves Claude Code's context compaction problem by maintaining state through YAML handoffs, ledgers, and a PostgreSQL-backed memory system that automatically extracts learnings from session thinking blocks. The plugin includes 109 skills, 32 specialized agents, 30 lifecycle hooks, and a 5-layer TLDR code analysis system that achieves 95% token savings, enabling complex multi-session workflows like TDD, refactoring, and formal theorem proving.

Key Features

109 meta-skills and workflow skills with natural language activation (/build, /fix, /tdd, /explore, /prove)
32 specialized AI agents for planning, debugging, implementation, and code review with orchestration patterns
30 lifecycle hooks that intercept SessionStart, PreToolUse, PostToolUse, and PreCompact events for context management
TLDR 5-layer code analysis (AST, call graph, CFG, DFG, program slicing) reducing token usage by 95%
PostgreSQL + pgvector memory system with automatic learning extraction from thinking blocks via background daemon
Continuity ledgers and YAML handoffs for preserving state across context clears and sessions
Formal verification support via Lean4 + Mathlib integration with 5-phase proof workflow
Built-in integrations with Braintrust (tracing), qlty (70+ linters), ast-grep, Perplexity, and GitHub search

Use Cases

  • 01Maintaining development context across multiple coding sessions without losing decisions and reasoning
  • 02Building complex features with guided workflows that orchestrate planning, validation, implementation, and testing agents
  • 03Debugging issues through specialized sleuth and profiler agents that perform root cause analysis
  • 04Performing safe refactoring with impact analysis, test-first implementation, and automated quality checks
  • 05Creating machine-verified mathematical proofs using Lean4 with compiler-in-the-loop feedback
  • 06Exploring unfamiliar codebases through semantic search and multi-layer code analysis with minimal token cost

Related Plugins

View more

andrej-karpathy-skills

A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.

Plugin216.1k

claude-mem

Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More

Plugin95k

Understand-Anything

Graphs that teach > graphs that impress. Turn any code into an interactive knowledge graph you can explore, search, and ask questions about. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and more.

Plugin84.9k

rtk

CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies

Plugin82.1k

Continuous-Claude-v3 — FAQ

What is Continuous-Claude-v3?+

Continuous-Claude-v3 is a Claude Code plugin system that adds persistent memory, multi-agent orchestration, and 109 workflow skills to Claude Code. It solves context loss on compaction through YAML handoffs, continuity ledgers, and a PostgreSQL memory system that automatically extracts learnings from sessions.

How do I install Continuous-Claude-v3?+

Clone the repository, navigate to the opc/ directory, and run the interactive setup wizard with 'uv run python -m scripts.setup.wizard'. The wizard handles Docker PostgreSQL setup, database migrations, and installation of all 109 skills, 32 agents, and 30 hooks into ~/.claude/.

Which AI clients does Continuous-Claude-v3 work with?+

Continuous-Claude-v3 is designed specifically for Claude Code CLI. It uses Claude Code's hook system, Task tool for agent spawning, and lifecycle events (SessionStart, PreToolUse, PostToolUse) that are unique to Claude Code.

What are the prerequisites and API keys needed?+

Required: Python 3.11+, uv package manager, Docker (for PostgreSQL), and Claude Code CLI. Optional API keys for enhanced features: Braintrust (session tracing), Perplexity (web search), and NIA (documentation search). Core features (continuity, TLDR, TDD workflows) work without any paid services.

Is Continuous-Claude-v3 free to use?+

Yes, Continuous-Claude-v3 is open-source under the MIT license and free to use. The core system works entirely with local tools and Docker PostgreSQL. Optional paid integrations (Perplexity, NIA, Braintrust) enhance functionality but are not required.

How does the memory system work?+

A background daemon monitors session heartbeats in PostgreSQL and spawns a headless Claude instance when a session ends. This daemon extracts learnings from thinking blocks and stores them in archival_memory with BGE embeddings. Future sessions automatically recall relevant learnings via the memory-awareness hook.

How do I install Continuous-Claude-v3?+

Open the source repository on GitHub and follow its README. Continuous-Claude-v3 is a plugin — MCP Agents Market links you directly to the official repo.

Is Continuous-Claude-v3 free?+

Continuous-Claude-v3 is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.

Related searches

Developer Tools PluginsProductivity & Workflow PluginsKnowledge & Memory PluginsBest MCP ServersBest SkillsBest AgentsTop rated listingsClaude Code toolsBrowse all categories