Draft

Not yet reviewed. Artifacts stay locked until you approve them.

AI Riders

Context is everything. But not everything belongs in context.

Your AI instructions are scattered across personal preferences, repository files, tool configs, and chat history — different owners, different scopes, no expiry. A Rider is the durable half: what every agent should assume about how you work, which technologies are policy rather than habit, and what an agent may do without asking. Compile it once, export it for Claude Code, Cursor, Copilot, and AGENTS.md.

Local-only. Nothing is sent anywhere.

Everything here runs in your browser. No account, no upload, no network request. Your draft is kept only in this tab unless you press Save draft to this browser, and anything that looks like a credential is redacted before it is stored or shown.

Policy domains
14

Areas a complete rider covers

Technology decisions
0/21

Capabilities with a chosen implementation

Actions governed
12

All at safe defaults

Sources imported
0

Instruction files you have pasted in

Start here

The interview is the fastest route to a first rider. Everything else refines it.

Guided interview — next release

The six areas

  • Audit

    Next release

    Paste what your agents load today and see what each rule actually is.

  • Optimize

    Next release

    Technology decisions and the actions an agent may take on your behalf.

  • Context Packs

    Next release

    Which policy loads for which task, and why it activated.

  • Policy Drift

    Next release

    Where instructions, packages, and tools stop matching the policy.

  • Reports

    Next release

    Exportable artifacts, context exposure, and validation coverage.

Founding waitlist

Keep this analysis. The workspace saves it.

The paid workspace adds saved reports and score history, API and CLI access, and team governance — opening to the founding waitlist first, at founding pricing locked in before public launch. This tool stays free either way.

Join the waitlist