Hidalgo
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
Hidalgo
A local page where a developer and an AI agent work through a task together: the agent shows its work, asks its questions, and waits for your verdict.
An agent's work is mostly decisions, and walls of text are a poor way to make them. Hidalgo turns that work into something to act on:
The agent's questions come as small choices you select, one at a time.
Each card or finding has its own conversation, so a discussion stays with the thing it is about.
The work itself takes shape where you can see it: the card being written, a review's findings, a diagram of the flow.
You spend your attention on thinking and deciding, not on reading.
Under the page is a local MCP server. The agent drives the page through its tools, and everything you do there comes back to the agent as events. The rules that matter are enforced by the server itself:
Approve stays blocked while a question is open.
A verdict counts only for the revision on screen.
A run doesn't close until every item has an outcome.
Using it with Claude Code
dist/ is committed, so a checkout runs with only Node 20 or later.
git clone git@github.com:J-Povea21/hidalgo.git
claude mcp add hidalgo --scope user -- node "$PWD/hidalgo/dist/server.mjs"Then add the Stop hook to ~/.claude/settings.json. It stops a turn from ending while a
run's page is waiting on an agent that isn't listening:
{
"hooks": {
"Stop": [
{
"hooks": [{ "type": "command", "command": "node /path/to/hidalgo/dist/stop-hook.mjs", "timeout": 5 }]
}
]
}
}The tools are open_session, add_items, ask, update_item, send_message,
create_diagram, wait_for_events, set_status, mark_done, list_runs and
close_session. A skill calls open_session first, shows its work with add_items and
ask, calls wait_for_events whenever it is the user's turn, and close_session at the
end.
Data lives in ~/.hidalgo/, or in ${CLAUDE_PLUGIN_DATA}/hidalgo/ when a Claude Code
plugin starts the server:
runs/<project>/<run id>/holds each run's state, event log and images;live/lists the pages open right now.
Closed runs are removed after 7 days; unfinished runs are kept until you decide.
Related MCP server: Kontrol
Developing it
pnpm install --frozen-lockfile # also turns on the pre-push check; no dependency runs code
pnpm run check # every check: what the pre-push hook runs
pnpm test # unit, component and end-to-end tests
pnpm run test:browser # the page in a real browser; run it before pushing
pnpm run format # Prettier
pnpm run build # rebuilds dist/; commit it with your change
pnpm run demo # the real server with a demo run, in your browserBefore changing it, read:
docs/conventions.md: the architecture, the code style, the security rules, and whydist/is committed;docs/testing.md: which kind of test to write, how a test reads, and how robust it has to be.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
MCP server for your apps' tools and custom tools, plus hosted AI agents and approval-gated workflows
MCP server for secureFlows: token-free URL builders and integration-linting tools for AI agents.
MarkupBase turns AI-generated Markdown and HTML into durable, versioned artifacts that people can review and discuss. Its MCP server lets agents publish new versions, preserve contextual comments, include hosted images, and respond to feedback through secure account-linked identities, creating a clear human review boundary without requiring real-time editing.
Impri MCP server — human-in-the-loop approval inbox for AI agents
Related MCP Servers
- AlicenseAqualityCmaintenanceAn MCP server that turns independent AI agents into a coordinated engineering team with shared task board, context, review loop, and enforced plan-implement-review-iterate workflow.24MIT
- AlicenseNot gradedqualityBmaintenanceA self-hosted MCP server that enables AI coding agents to read, edit, search, and run code in local projects with human review loops and policy controls.MIT
- AlicenseNot gradedqualityDmaintenanceAn MCP server that lets ChatGPT or any MCP client securely delegate coding tasks to a local Claude Code instance, with git checkpointing, approval gates, and structured results. Supports code review, test running, and rollback via simple tool calls.14 npmMIT

mcp-reviewofficial
AlicenseAqualityBmaintenanceAn MCP server for in-loop design review of web previews. It enables agents to submit a preview URL, receive structured findings with suggested fixes, and recheck after applying changes, while never editing code itself.52MIT