Agentpack

Created By
ihorponom7 days ago
Agentpack is a local-first continuity layer for AI coding agents. It stores reviewed task state inside a repo-local .agentpack/ ledger so Codex, Claude Code, Cursor, Claude Desktop, and other MCP clients can resume work without rebuilding context from chat history. Agents can record durable decisions, failed approaches, evidence, reviewed source conclusions, task passports, and checkpoints, then load compact resume context under a token budget. Git keeps the code history; Agentpack keeps the task state a future agent needs to continue safely. Agentpack runs locally, exposes a stdio MCP server, has zero runtime dependencies, makes no network calls during normal operation, and is designed for repo-scoped coding workflows rather than general chat memory.
Overview

Agentpack

Repo-native task continuity for MCP-connected coding agents.

Agentpack gives Codex, Claude Code, Cursor, Claude Desktop, and other MCP clients a local repo-scoped task ledger. Agents can record decisions, dead ends, evidence, source conclusions, task passports, checkpoints, and compact resume context so future sessions can continue without rebuilding context from chat history.

What it helps with

  • Continuing after chat compaction or a fresh agent session
  • Switching between Codex, Claude Code, Cursor, and Claude Desktop
  • Recording durable task decisions and failed approaches
  • Preserving reviewed source conclusions with file hashes
  • Creating checkpoints and handoffs for repo-scoped coding work

Install

npm install -g agentpack-cli
agentpack init
agentpack install codex --write
# or: agentpack install claude --write
# or: agentpack install cursor --write
agentpack doctor

Server Config

{
  "mcpServers": {
    "agentpack": {
      "command": "npx",
      "args": [
        "-y",
        "agentpack-cli@latest",
        "mcp",
        "--root",
        "/absolute/path/to/your/repo"
      ],
      "env": {
        "AGENTPACK_ROOT": "/absolute/path/to/your/repo"
      }
    }
  }
}

Recommend Servers

View All
Bring your real authenticated browser session to AI coding agents. Local-first MCP server + Chrome MV3 extension. No cloud. No telemetry.
@Cubenest

peek records the user's actual logged-in browser (DOM via rrweb, console events, network metadata, optional response bodies via opt-in Deep capture) through a Chrome MV3 extension. The extension ships events through a native-messaging stdio bridge to a local MCP server (peek-mcp), which persists them to a SQLite database at ~/.peek/sessions.db. AI coding agents (Claude Code, Cursor, Cline, Windsurf) read sessions from the database via 10 MCP tools: Tool What it does list_recent_sessions List recently recorded sessions (id, origin, ts, event count). get_session_summary LLM-readable narrative summary of a session. get_session_console_errors Console errors recorded in a session. get_session_network_errors Failed/notable network requests in a session. get_user_action_before_error Last N user actions before a console error. generate_playwright_repro Generate a runnable Playwright test from a session. get_dom_snapshot Reconstruct the DOM at a given timestamp. query_dom_history Timeline of attribute/text changes for a selector. request_authorization Side-panel consent for write actions (Level 3). execute_action Dispatch a UI action (gated by permission level + destructive blocklist). Why local-first matters Every other "browser session for AI" tool ships to a vendor cloud. peek's SQLite + extension live on the user's machine — no remote endpoints, no telemetry. The privacy policy (docs/peek/PRIVACY_POLICY.md) is the source of truth. Install # 1. Add the MCP server to Claude Code claude mcp add peek -- npx -y @peekdev/mcp # 2. Install the Chrome extension from the Chrome Web Store # (link added once the CWS listing is approved)

2 days ago
Voyei

4 hours ago