125 AI-ready tools, semantic wiki search, and Claude Code skills - Monolynx isn't just another dashboard. It's a knowledge base your AI agent works with as if it were a team member.
Twelve integrated modules that cover the full project lifecycle.
Error tracking with smart fingerprinting. Catch exceptions from your apps with a lightweight SDK.
Backlog, Kanban board, sprints, and story points. Everything for agile project management.
URL health checks with uptime tracking, response time history, and instant alerts.
Cron job monitoring with dead man's switch. Get alerted when your scheduled tasks stop running.
Markdown documentation with page hierarchy, image uploads, and AI-powered semantic search.
Interactive dependency graph powered by Neo4j. Visualize how your code modules relate to each other.
Cross-project billing with status workflow (draft → sent → paid), ticket freezing, and attachment management.
Cross-project work reports with multi-select filters, date ranges, and PDF export for stakeholders.
Personal cross-project scheduling on a Gantt chart and monthly calendar, independent of sprint assignment.
Observability for AI agent work - pipeline, step and job hierarchy with logs, durations and upward status propagation.
Command-line client (monolynx, alias mnx) with the same operations as MCP and pipeline-ready JSON output.
Public blog with ready-made blocks, a markdown twin for AI models, structured data, feeds and a lint before publication.
Give your AI assistant direct access to Monolynx: add it to Claude.ai as a custom connector over HTTPS, or wire it into any client that speaks the Model Context Protocol. Manage tickets, search your wiki, and query your dependency graph in natural language.
Are you an AI agent? Start with the guide for AI assistants or llms.txt.
Install Monolynx as a native Claude Code plugin. Skills and agents show up as /monolynx:* commands.
Install
/plugin marketplace add https://gitlab.com/piotrkrych/monolynx.git
/plugin install monolynx@monolynx
In the plugin config set your token (mcp_token, osk_... format), optionally mcp_endpoint (defaults to https://monolynx.com/mcp) and project_slug.
Available commands
/monolynx:setup
/monolynx:next
/monolynx:brief
/monolynx:work
/monolynx:work-simple
/monolynx:resume
/monolynx:ticket-create
/monolynx:ticket-review
/monolynx:sprint-end
/monolynx:sprint-run
/monolynx:mr-queue
/monolynx:retro
/monolynx:search
/monolynx:help
Add Monolynx directly in Claude.ai (web or desktop) as a custom connector over HTTPS. No install, no config files.
Connector URL
https://your-instance/mcp
Replace your-instance with your Monolynx instance host.
For Claude Code or any MCP client, add an entry to your .mcp.json:
{
"mcpServers": {
"monolynx": {
"type": "http",
"url": "https://your-instance/mcp",
"headers": {
"Authorization": "Bearer YOUR_TOKEN"
}
}
}
}The flagship AI tool install_monolynx_skills drops all 24 ticket and wiki skills into any project connected to Monolynx - Claude Code, Codex or Cursor. No ZIPs, no manual placeholder replacement.
Just say it in Claude
Install the default Monolynx skills for project acme-inc.
Agent wywoła MCP i zapisze skille do katalogu swojego runtime'u:
.claude/skills/,
.codex/skills/ lub
.cursor/skills/.
You get
Monolynx is free and open source. Self-host it on your own infrastructure, customize it to your needs, and keep full ownership of your data.
View on GitLab