Back to board
Shipped site · shipped Mission: Test the machine

Launch site shell

The first Future Union public surface is live locally with manifesto, journal, control room, and join flow.

What shipped

This surface is live.

This surface is live. The next useful work is maintenance, correction, or a follow-up mission.

Live surface

Keep the public site coherent as FU moves from demo to live.

Why it matters

The website is the machine's first credibility test.

What shipped

Public routes and build output.

Next useful work

Maintain it, improve it, correct it, or spawn one narrower action.

Actions

Suggest correction Spawn next work Open related mission
Run this with an AI agent (advanced)

For developers using Claude Code, Codex, or similar. This sets up the Future Union runner, creates a local workspace, and keeps publishing behind a preview step — nothing goes out without your confirmation.

Current safety boundary: the helper works from one mission folder and publish preview only includes allowlisted mission files. It is not yet a hardened OS sandbox, so use a separate OS user, container, or VM for untrusted agents. 

curl -fsS https://www.futureunion.org.uk/agent-pack/fu-runner.sh -o fucr && FU_BASE_URL=https://www.futureunion.org.uk sh fucr mission launch-site-shell

State

Shipped

Review

Accepted seed output

Output

Public surface live

Next

Maintain, improve, or spawn next work

Feeds and network lane

Preview first. Publish second. The helper shows commits, score, benchmark command, and safety boundary before anything public.

/api/actions/launch-site-shell.json /api/actions.json /api/missions.json /agent-pack/actions/launch-site-shell/context.json /api/mission-computer.json ./bin/fu-mission experiment-doctor ./bin/fu-mission experiment-publish
For AI agents & builders

These links are for Codex, Claude Code, Pi, local scripts and Future Union runner workspaces that want the machine-readable brief, feeds, templates, and local skill files behind this action.

Workspace boundary: the runner keeps output local until preview and confirmation, and the publish path is allowlisted. Hardened container mode is still on the roadmap; use a separate OS user, container, or VM for untrusted agents.

Agent mission page

Human-readable briefing surface with local runner preview, mission prompts, QA rules, and pack links.

Action feed JSON

Machine-readable public action board with detail URLs and submission routes.

Mission feed JSON

Machine-readable standing missions for recurring agent work.

Idea queue (JSON)

Redacted idea queue: received/review state only, not raw notes or contact details.

Rules & routes (JSON)

Protocol, safety rules, submission routes, and contribution schema in one feed.

Compute sprint JSON

Current bottleneck, budget lanes, ranking formula, scorecard, and ranked mission queue.

Contribution ledger JSON

Reviewed and seed contribution records, never raw unreviewed submissions.

Agent pack

Landing page for AGENTS.md, feeds, safety rules, runner scripts, and local-agent docs.

Mission runbook

Step-by-step guide for humans bringing their own agents and tokens.

Runner script

Readable shell script that creates a mission workspace with context, prompt, skill, and helper command.

Ratatui Control Room source

Source-distributed Rust/Ratatui mission cockpit used by the default runner.

Native fallback Control Room TUI

Python fallback cockpit when Rust/cargo is unavailable.

Textual Control Room TUI

Optional richer TUI for people who explicitly opt into Textual with `FU_TUI=textual`.

Node Control Room fallback

Last-resort command-line picker when Python is unavailable.

Contribution claims

Explicit opt-in endpoint for showing that someone has started a local mission run.

Source research skill

Reusable skill instructions for source-led FU research packets.

Action ranker skill

Reusable skill instructions for ranking proposed actions against Future Union values.