AI coding agents produce great documentation, but most of it gets thrown away. Its easy enough to save it to markdown files, or push it to a document store via MCP but raw documentation gets out of date very fast, becoming hard to use and misleading.
Telos uses AI agents to manage documentation for you. Its the best documentation you've ever had, that you'll never read. That's because our AI agents will also read it for you and work with your dev agent to provide relevant context during development and support work.
In Telos, all dev work and documentation is stored in tickets and can be accessed via our web portal and/or using MCP tools.
Tickets are processed and turned into a living blueprint by our AI agents.
Telos allows you to manage multiple applications and keep them silo'ed.
After you signup, create an application.
You will need a personal access token to use our MCP server. Create one from profile menu by selecting 'Edit profile', then select 'Personal Access Tokens'
Add the Telos MCP server to your development AI agent, such as Cursor, Claude or OpenAI.
"mcpServers": {
"telos": {
"type": "sse",
"url": "https://go.telosready.com/mcp?apiKey=personal_access_token"
}
}
The blueprint is the source of truth for your application. It is designed to keep track of the business context that exists outside of your codebase.
The blueprint is maintained by our AI agents.
Resources
Resources are the important and distinct "moving parts" used to host and run your application. Resources typically represent infrastructure, integrations or key services relied on by the app.
Concepts
Concepts are the ideas and definitions that are specific to your business and application. They don't describe how the application works, they describe how the business works.
Stories
Stories are what the software does and the jobs that people (and AI agents) use the software for. As these stories change over time, the blueprint will be kept up to date.
Start using Telos in minutes using our MCP tools. Tools that rely on AI agents only work with paid plans. Ask us if you'd like a trial.
Get Ticket Details
Gets the full context for a specific ticket. This can include comments, transcribed videos, screenshots and documentation. A ticket can include a development task, an action list and/or documentation. Can be used with a simple prompt: "Implement ABC123"
Get Next Task
Our AI agent will analyse open tickets and identify what the next task is to be completed. A task is the next step of a ticket, based on checklists contained with the ticket. Tasks can be assigned to multiple users - such as AI for coding, human for review.
Requires a paid plan.
Ask Question
Our AI agent will review all tickets and look for relevant concepts, resources and stories in the application blueprint and answer the question. The quality of all answers is tracked, so you can look for information gaps.
Requires a paid plan.
Get Git Branch For Ticket
Helps git branches to be managed across different agents and enables the dev agent and team to commit code to the right branches.
Add Ticket Comment
When working on a specific ticket, documentation can be added directly to the ticket as a comment. This provide the best context for the documentation.
Add Documentation
When a specific ticket isn't provided, the dev agent can add documentation and it will be saved as a knowledge ticket.
Knowledge tickets are only processed by our AI agent with a paid plan.
Mark Ticket As Blocked
When a ticket requires further information or the AI agent can't proceed for any reason, it can mark a ticket as blocked and provide a reason why. This will remove the ticket from the workflow until it is unblocked.
Update Action Item
Tickets can contain action lists (using markdown checklists) and the AI dev agent is able to mark action items as in-progress and completed.
We love solving business problems using AI-powered software, for founders who are relentless about improving their systems so they can stay ahead of their client's ever changing demands.