Controlled execution for AI coding agents

Ship the useful part of AI. Keep the boundary.

SafeShip turns an approved GitHub task into a constrained proposal, an isolated verification run, and a draft pull request your team can actually review.

Six-minute product walkthroughSafeShip founding evaluation
ScopedPaths, commands, and acceptance criteria declared up front.
SandboxedNon-root, network-disabled verification with bounded evidence.
ReviewableBase commit and proposal digest bound before draft publication.
The workflow

From approved issue to evidence-backed draft PR.

SafeShip makes the agent’s work legible. Every run has a boundary, a result, and a deliberate human gate before anything reaches GitHub.

01 / APPROVE

Declare scope

Choose the repository, allowed paths, checks, and acceptance criteria.

02 / PROPOSE

Generate files

The model returns structured file contents. It never writes directly to the worktree.

03 / VERIFY

Run isolated

Apply the proposal inside a disposable, non-root Docker workspace.

04 / REVIEW

Publish draft

Recheck the binding, then create a draft PR only when a human chooses publish.

Built for trust

A green test is not the whole story.

SafeShip records the changed paths, command output, sandbox state, base commit, proposal digest, and acceptance criteria so your team can review what actually happened.

  • Repository and user allowlists
  • Network-disabled container runner
  • Bounded output and resource limits
  • Idempotent submissions
  • Base-commit and proposal-digest rechecks
Evaluation licence

Free for 30 days. Open to anyone.

Evaluation access is free, unvetted, and granted the same day. Email us, get read access to the evaluation repository, and try it against a disposable repository of your own. Continue to a paid setup only when you decide it belongs in your process.

Request evaluation access
Free for 30 days
  • No payment details, no waiting list
  • Read access to the signed release and source
  • Confidential — do not redistribute, see the licence
  • Contributions welcome, gated on a CLA before merge
  • Report a broken safety claim directly to us
After the evaluation

$750 setup, then $250/month.

Setup is waived for the first 100 customers — use code FIRST100 at checkout. No long-term contract; cancel any time.

Continue to checkout
$750 setup + $250 / month
  • Setup fee waived for the first 100 customers (code FIRST100)
  • Policy and GitHub App setup
  • Ongoing verified draft-PR workflow
  • Cancel any time
Request access

Someone else wants to try it?

Send them here, not your clone. Anyone can email safeship@leadrescue.us and get their own access the same day — there is no qualification to meet.

Full terms are in the evaluation licence agreement. Every release is checksummed and signed — see verifying a release before you install anything.