10 min read
Attach an MCP server
Bring your own tools. Point Helios at any MCP server, choose which of its tools agents may call, and they become part of the agent's reach.
- 01
Add the server URL
From Settings → MCP Servers, add a server by its streamable HTTP endpoint. Helios validates the URL against its SSRF protections before it connects, so an attached server cannot be pointed at your internal network.
- 02
Choose how it authenticates
Servers connect with no auth, a bearer token, custom headers, or an OAuth flow. Tokens and headers are stored envelope-encrypted and attached server-side at call time, the same way integration credentials are.
- 03
Review the tools it exposes
Helios lists every tool the server advertises, with its description and input schema. Read the list before you attach it to anything — this is the surface you are handing your agents.
- 04
Filter down to the tools you want
A tool filter narrows the server to a named subset. A server that exposes forty tools can contribute the three that matter to a task, which keeps the agent focused and the blast radius small.
- 05
Attach it to a workflow and run it
Enable the server on a workflow, name the tools in the prompt, and run it manually. The run record shows each MCP tool call the agent made alongside its integration calls, so a mixed run reads as one story.
- 06
Point another agent at Helios
MCP runs both directions. Helios exposes its own API over MCP, so an agent living somewhere else searches the Helios API surface for the endpoint it needs and calls it — listing projects, creating a workflow, reading its runs. The same credentials and the same roles apply from that side.
Keep going
Next in the track
10 min
Run your first agent
Describe a task in plain English, give the agent a tool to reach for, and watch it work. Ten minutes from an empty workspace to a run you can read end to end.
8 min
Use secrets and variables
Keep the values a task needs out of the prompt. Variables hold configuration, secrets hold credentials, and both are scoped to the level where they belong.
7 min
Pin repositories as context
Give an agent standing context: pin a GitHub repository and the agent carries it into every run without being told again.
Templates
Put it to work
Templates that use what this guide covers.
Engineering
Pull request reviewer
Reviews every pull request against your conventions, leaves inline comments, and posts a weekly digest of review load and CI health.
Engineering
Incident postmortem drafter
Reconstructs the incident timeline from Slack and GitHub, names the triggering change, and drafts the write-up in Notion.
Operations
Nightly data quality checks
Runs your invariants against production every night and tells the on-call analyst which rows broke and since when.
Try it in your workspace.
Start free, connect a tool, and describe the first task in plain English.