Now onboarding first customers
Ship to production with your AI agent.
Without the 3am Slack from your AI agent.
Same git-push-to-deployed as Railway. Half the price. Guardrails so Claude or Cursor can ship without dropping your prod database.
Built so your AI agent cannot break things
Existing tools were built for humans typing carefully. Fieldwick is built for humans plus AI typing fast - same git push, same auto-TLS, different threat model.
Soft delete by default
Every destructive action - delete service, drop database, remove domain - is reversible for 7 days. Your AI agent can mistake-fix instead of you restoring from backup at midnight.
Confirm tokens for risky ops
Production deletes require a fresh nonce token from the dashboard. An agent reading your codebase can suggest the operation but cannot execute without a human-in-the-loop step.
Spend caps you set
Set a monthly cap. We throttle deploys and scale-ups before the cap is hit, not bill you $4,000 because your agent left a debug loop running over the weekend.
MCP server, native
Drop our config into Claude Code or Cursor in 30 seconds. Your agent lists services, deploys, tails logs, sets env vars through the same scoped, audit-logged API as the dashboard.
Per-token audit log
Every API call records who, what, when - separately for human users vs API tokens vs agent sessions. Sortable, exportable, queryable. Your agent does not get to deny what it did.
Receipts: every mutation audited
Every confirm-token consume and every service change (create, deploy, env_set, scale, restore, delete) writes an audit log row. Read them back via the API, the CLI, or let your agent check via the MCP list_audit_log tool. Your agent reads its own past actions back through the same scoped, audit-logged surface you do.
Half the price. Built for the AI era.
Same git-push-to-deployed workflow. Same Nixpacks. Same custom domains with auto-TLS. Half the meter rates plus the guardrails your AI teammate needs.
The price is structural, not a promo: we run on dedicated bare metal with real cores, NVMe SSDs, and a 10 Gbps private network. No hypervisor tax, no reseller markup stacked on another reseller.
| Railway | Fieldwick | |
|---|---|---|
| Pricing | ||
| Memory | $0.010 / GB-hour | $0.005 / GB-hour |
| CPU | $0.020 / vCPU-hour | $0.010 / vCPU-hour |
| Egress | $0.10 / GB | $0.05 / GB |
| Storage | $0.25 / GB-month | $0.10 / GB-month |
| Free credit | $5 / month | $5 / month |
| Workflow + AI safety | ||
| Build system | Nixpacks | Nixpacks (drop-in) |
| Custom domains | Auto-TLS | Auto-TLS |
| Soft delete by default | No | 7-day recovery |
| Spend cap throttling | No | Built-in |
| Confirm tokens on destructive ops | No | Required |
| Per-token audit log | No | Sortable + exportable |
| AI-agent MCP server | No | First-class |
Pricing reflects published rates as of April 2026. Free credit applies to net usage after each billing cycle resets.
Hit the API right now
Two real endpoint shapes - the same JSON your dashboard, CLI, and agent read. Click to fire a request and see the response.
Fixture mode: responses are canned but match the shape of the real api.
# click an endpoint to fire a request
Dashboard
The AI-safety rails, visible on every page.
Not hidden behind "enterprise only" checkboxes. Every team gets the same scoped api, the same confirm-token guard, the same spend cap gauge, the same audit log. Agents and humans use identical surfaces.
- Cap
- $100 / month
- Action at cap
- throttle
- This month
- $82.00
Deploys refuse with HTTP 402 when projected spend would push you past the cap. Agents hit the same wall humans do.
Soft-delete is reversible for 7 days via the Restore button. After that, the sweeper removes the record + sealed env vars permanently.
Every destructive op - delete, drop, remove - requires a token from the dashboard OR an explicit CLI flag. Humans in the loop by default; agents can suggest, not execute.
- service.deployactor: cursor · 2m ago
- service.env_setactor: you · 12m ago
- confirm_token.consumedactor: you · 1h ago
- billing.invoice_paidactor: system · 2d ago
Every mutation writes a row: human via dashboard, human via CLI, agent via api token, system via webhook. Same table, same queryable surface.
Shown: real dashboard components (iters 155-163) rendered inline. The live build runs at localhost:3001 today and moves to a hosted URL once auth ships.
See it running
Demo recording in progress. Here is the minute-by-minute script of what the video will walk through.
- 1
Deploy end-to-end in the docker stack
`pnpm dev:test:api` drives the full customer flow: auth, create team/project/service, seal env var, trigger deploy, edge-served HTTP response. ~30 seconds on a laptop.
- 2
Tail live logs through the CLI
`fieldwick logs <service-id> --follow` pipes Nomad alloc stdout/stderr through Fastify chunked transfer; ctrl-C cancels cleanly through to Nomad (no leaked tails).
- 3
Talk to your deploys from Claude Code
The MCP server (apps/mcp) exposes list_services + get_service + tail_logs as first-class tools. Ask Claude "tail my api service" - it calls the MCP tool, shows you the log buffer, explains what it sees.
- 4
Try to delete a service as an agent
DELETE without a confirm token returns 428 Precondition Required. The agent reads the error, prompts the human to authorize, the human clicks through a dashboard modal that issues a 60-second token, then the DELETE succeeds. Human-in-the-loop enforced by the protocol.
Pay-as-you-go at half Railway rates
Every tier includes the AI-safety guardrails. The difference is seats, retention, and how fast we get back to you when something burns.
Free
For side projects + experiments. No card up front.
- 1 service, 1 region
- Soft delete + 7-day log retention
- Community Discord support
- Spend cap enforced at $0 - we throttle, never bill
Pro
For builders shipping with AI. Half Railway meter rates.
- Unlimited services across both regions
- All AI-safety guardrails (confirm tokens, audit log, MCP server)
- 30-day log retention
- Spend cap throttling at the limit you set
- Email support, 24h response
Team
For teams with shared deploys + compliance needs.
- Everything in Pro
- Multi-seat with role-based access
- Audit log export to S3-compatible storage
- Custom SSO via your identity provider
- Priority support, 4h response
PAYG meter rates: $0.005 / GB-hour memory, $0.010 / vCPU-hour CPU, $0.05 / GB egress, $0.10 / GB-month storage. Same on every tier.
Get early access
We're onboarding first customers this week. Drop your email and we'll set up a 15-minute walkthrough.
We'll only email you about Fieldwick. No newsletter. No marketing drip. Unsubscribe anytime.