Описание
Time-aware work sessions for AI agents.
README
Owlyn MCP
Owlyn MCP is a local-first MCP work-session supervisor for AI coding agents that need deadlines, checkpoints and structured continuation decisions instead of a simple timer.
It is for developers using agentic coding tools who want long-running work sessions to remain explicit, inspectable and bounded by safety rules.
AI agents often stop too early. You ask an agent to work until 06:00, it finishes the first task in 20 minutes, then says "done." Owlyn exists to make the continuation decision explicit.
Preview
Screenshots / GIF demo coming soon.
The repository includes the Owlyn logo above and example session documentation in docs/EXAMPLE_SESSION.md.
Features
- Stdio MCP server built with the official MCP SDK.
- Local SQLite session storage through
better-sqlite3. - Work-session lifecycle tools: start, status, checkpoint, should-continue, plan-next, end, list-sessions and report.
- Deadline and timezone handling for time-aware work sessions.
- Conservative continuation policy that stops for high-risk, destructive or approval-requiring work.
- Host setup and agent instruction docs for integrating Owlyn into coding-agent workflows.
- MCP smoke test that exercises the core tool flow.
Installation
Requirements:
- Node.js 20+
- npm
git clone https://github.com/btitkin/OwlynMCP.git
cd OwlynMCP
npm install
npm run build
Installation instructions are based on the current repository structure and should be verified for each MCP host.
Quick Start
Run the stdio MCP server:
node dist/index.js
Generic MCP server config:
{
"mcpServers": {
"owlyn": {
"command": "node",
"args": ["/absolute/path/to/OwlynMCP/dist/index.js"],
"env": {
"OWLYN_DB_PATH": "/absolute/path/to/.owlyn/owlyn.sqlite"
}
}
}
}
More host setup notes are in docs/HOST_SETUP.md.
Examples / Use Cases
- Start a bounded AI-agent work session with a goal and deadline.
- Save checkpoints that include completed work, next tasks, validation and blockers.
- Ask whether the agent should continue after the current task finishes.
- Rank safe next tasks when time remains in a session.
- Produce a session report before ending work.
Roadmap
See docs/ROADMAP.md for the release path from alpha to v1.0.
Future ideas after v1.0 include optional JSON/Markdown export, optional session templates, host-specific config helpers, per-project session profiles, and optional pause/resume controls.
These items are not implemented in the current alpha release.
Status
Alpha: the current public release is v0.1.0-alpha.2, includes tests, documentation, GitHub Actions CI and a manual MCP smoke test, and has been validated in Codex MCP host on Windows. Broader host compatibility should still be treated as pending unless verified in the target host.
Codex-first status
Owlyn MCP is currently validated in Codex MCP host on Windows.
Extended Codex validation scenarios passed for continuation decisions, safety gates, focused mode, marathon mode, deadline handling, task ranking, report, end, restart/reload behavior, and persistent configured OWLYN_DB_PATH behavior. Clean temporary DB path validation remains optional/pending unless separately tested.
The next focus is deeper Codex workflow validation:
- long-running work sessions
- continuation decisions
- checkpoint quality
- clean temporary database-path validation
- safe next-task planning
Other MCP hosts remain pending.
Codex docs:
General Host Validation Status
Validated:
- Codex MCP host on Windows
Pending:
- Cursor
- Claude Desktop
- other MCP stdio hosts
See real host validation, host setup, and the shared validation prompt.
More Docs
- FAQ
- Host setup
- Real host validation
- Codex MCP setup
- Codex agent instructions
- Codex validation scenarios
- Codex validation log
- Validation prompt
- npm publishing preparation
- Roadmap
- Tool reference
- Agent instructions
- Social post draft
Development
npm install
npm run typecheck
npm run build
npm test
npm run smoke:mcp
npm run check
Release checklist: RELEASE_CHECKLIST.md.
License
MIT. See LICENSE.
Установка Owlyn
У этого сервера нет опубликованного пакета — он собирается из исходников. Открой репозиторий и следуй инструкции в README.
▸ github.com/btitkin/OwlynMCPFAQ
Owlyn MCP бесплатный?
Да, Owlyn MCP бесплатный — установка в пару кликов через Unyly без оплаты.
Нужен ли API-ключ для Owlyn?
Нет, Owlyn работает без API-ключей и переменных окружения.
Owlyn — hosted или self-hosted?
Self-hosted: сервер запускается локально на твоей машине командой из раздела установки.
Как установить Owlyn в Claude Desktop, Claude Code или Cursor?
Открой Owlyn на unyly.org, выбери вкладку своего клиента (Claude Desktop, Claude Code, Cursor) и нажми Install — конфиг сгенерируется автоматически, без правки JSON.
Похожие MCP
Fetch
Web content fetching and conversion for efficient LLM usage.
AWS KB Retrieval
Retrieval from AWS Knowledge Base using Bedrock Agent Runtime.
автор: modelcontextprotocolSpring AI MCP Server
Provides auto-configuration for setting up an MCP server in Spring Boot applications.
llm-analysis-assistant
A very streamlined mcp client that supports calling and monitoring stdio/sse/streamableHttp, and can also view request responses through the /logs page. It also
автор: xuzexin-hzCompare Owlyn with
Не уверен что выбрать?
Найди свой стек за 60 секунд
Автор?
Embed-бейдж для README
Похожее
Все в категории ai
