Human-in-the-loop for Continue
Continue is the open-source coding assistant for VS Code and JetBrains, configured through a single config.yaml. Add Orako under mcpServers and Continue's agent mode can ask the domain owner when a decision needs a human, then keep working with the reply.
Connect Orako to Continue
- 01
Add the mcpServers block below to config.yaml (or through the Continue UI).
- 02
Switch to agent mode and approve access in the browser on first use:
askandsearch_historyjoin the tool list. - 03
Human-only decisions route through Orako; the session resumes on the answer.
mcpServers:
- name: Orako
command: npx
args: ["-y", "mcp-remote", "https://app.orako.io/mcp"]MCP tools run in Continue's agent mode only; switch modes before testing. The mcp-remote proxy opens a browser tab for OAuth the first time — nothing to store in config or Continue's secret store.
New to MCP in Continue? See Continue’s docs.
What Continue can ask
Continue's agent mode, with your team on the tool list.
Requirements that live in someone's head
The spec says "like the old export but better." Continue asks the PM what better means, gets a reply, and builds that.
JetBrains and VS Code, same loop
Continue runs in both IDE families with the same config.yaml, so one Orako entry gives every editor in the team the same ask-a-human tool.
Onboarding questions that stop repeating
New contributors' agents ask the same setup questions. After the first resolved conversation, search_history can find it instead of pinging the maintainer again.
Continue + Orako, answered
Does Continue support MCP natively?
Yes. MCP servers are a top-level mcpServers list in config.yaml, each with a name, command, and args. Continue exposes those tools in agent mode, which is where an ask-a-human tool belongs.
Why don't I see Orako's tools in chat mode?
By design: Continue only runs MCP tools in agent mode. Switch the mode selector to agent and the Orako tools appear. Chat and edit modes never call tools, so they can't ask anyone anything.
Do my teammates need this agent installed to answer?
No. Only the developer running the agent connects Orako. The people answering receive each question in Slack, Discord, or the Orako dashboard and reply in the thread. They never touch an editor, a terminal, or an MCP config.
What happens to the answers?
Every resolved conversation becomes searchable team history with its authors, date, tags and full context. Agents call search_history before asking, so a question your team already answered can be reused instead of pinging anyone again.
Put a human in the loop for Continue
Give Continue the one thing it can’t generate: the answer only your team knows.