The local Shipmoor command line

One CLI for the complete agent verification loop.

Check whether an agent completed the task it was given. Verify that tests really ran and were not weakened. Scan for structural defects. Run advisory review with your own agent. Keep the same checks inside the agent loop and continuous integration.

Shipmoor CLI is the local interface for the complete agent verification loop. The free Community Scan catches structural defects. Shipmoor IC adds Claim Check, Test Evidence, advisory Code Review, Agent Skills, and the Agent Harness.

Claim Check is the main verification command. It decides whether an agent generated change earned its completion claim. The other commands provide the evidence and workflow controls that make that verdict trustworthy.

  • Runs locally
  • No source upload
  • No Shipmoor hosted model
  • Claim Check
  • Test Evidence
  • Deterministic Scan
  • Advisory Code Review
  • Agent Skills
  • Agent Harness
$ shipmoor claim-check --intent-prompt "add retry handling to failed webhooks" --agent claude . Claim Check is BLOCKED. One obligation has diverged.  The build passed. The tests passed. The scan passed.  AC 02 is blocked because retry stops after the first failure.  The evidence is located in src/webhooks/retry.ts at line 47.  The fix packet asks for a check that proves retries continue to the configured limit. $ shipmoor claim-check --agent claude . Claim Check is READY. Both obligations are verified.The attestation was written to .shipmoor/claim-check.vsa.json.

Claim Check is the main verification command. It returns a verdict and gives the agent a concrete path to resolve every blocker.

Start with the free Scan

The Community Scan works immediately after installation and does not require an account. Run it in the repository where your coding agent made the change. It checks the changed files for high confidence structural defects before you push.

shipmoor scan --changed

Use it after Cursor, Claude Code, Codex, Copilot, Aider, or any coding agent that changes files in your repository.

  1. Let the coding agent finish its change.
  2. Run shipmoor scan on the changed files.
  3. Resolve the reported structural defects.
  4. Run the scan again before you push.

Six command groups cover the full loop

Scan is the free structural layer. Shipmoor IC adds the commands that verify intent, test credibility, reading level risk, and the agent workflow itself.

  • Verify the change against the original task

    shipmoor claim-check --intent-prompt "add retry handling" --agent claude .
  • Verify that tests ran and were not reduced or weakened

    shipmoor test-evidence .
  • Scan changed files for deterministic structural defects

    shipmoor scan --changed
  • Run advisory review with your own coding agent

    shipmoor review --agent claude --from main --to HEAD
  • Install Shipmoor skills into your coding agent

    shipmoor skills install claude
  • Route verification results back into the agent loop

    shipmoor harness mode feedback

Claim Check owns the final completion verdict. The other commands produce evidence, provide advisory findings, or keep verification active while the agent works.

One command with four levels of access

There is one installed shipmoor command. Community Scan works without login. Shipmoor IC unlocks the complete individual verification loop. Team and Enterprise plans will add shared enforcement and governance.

  • Free

    Community

    Local structural scans for every developer.

    • Changed, staged, diff, and patch scan modes
    • Human, JSON, and SARIF output
    • Stable exit codes for continuous integration
    • No account required
  • 19 dollars per month

    Shipmoor IC

    The complete local verification loop for one developer.

    • Claim Check
    • Test Evidence
    • Advisory Code Review
    • Agent Skills and Agent Harness
    • Advanced command line and editor features
  • Coming soon

    Team

    Shared enforcement for engineering teams.

    • Everything in Shipmoor IC
    • Managed continuous integration gates
    • Shared policy and baselines
    • Pull request comments
  • Coming soon

    Enterprise

    Governance and deployment controls for larger organizations.

    • Everything in Team
    • Organization administration
    • Single sign on and audit logs
    • Private or self hosted runners

Unlock Shipmoor IC with one login

Install the CLI once. When you want Claim Check and the rest of the individual verification loop, sign in from the same command. Your local entitlement unlocks the paid commands.

  • shipmoor login
  • shipmoor whoami
  • shipmoor capabilities --json

There is no second binary and no separate login for the editor or agent harness.

Local by design

The CLI runs in your development environment or continuous integration runner. Shipmoor does not upload your source code and does not host or call a model. Features that need an agent use the coding agent and provider relationship that you already control. The deterministic Claim Check floor can run without a model.

Use the free Scan in continuous integration

The same deterministic Scan that runs before a push can gate a pull request and publish SARIF findings. Claim Check projects can also run the model free floor against their committed acceptance contract.

name: Shipmoor scan

on:
  pull_request:

jobs:
  shipmoor:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0

      - name: Install Shipmoor
        run: curl -fsSL https://dl.shipmoor.dev/install.sh | bash

      - name: Run the deterministic Scan
        run: shipmoor scan --diff origin/main...HEAD --fail-on high --sarif --output shipmoor.sarif

      - name: Upload SARIF
        uses: github/codeql-action/upload-sarif@v3
        with:
          sarif_file: shipmoor.sarif
        if: always()

For an approved Claim Check contract, shipmoor claim-check --floor-only runs the reproducible verdict path without a model or model secret.

From agent output to an earned verdict

Shipmoor belongs inside the coding loop, before a completion claim reaches human review. The CLI collects evidence early and gives the agent a direct path to resolve every blocker.

  1. The agent changes code then
  2. Shipmoor collects evidence then
  3. Claim Check returns a verdict then
  4. A reviewer evaluates the change then
  5. The change can merge

Human review still matters. Shipmoor makes sure that review starts with an explicit task, credible evidence, and an honest completion state.

Install the CLI and start locally

Use the Community Scan for free. Start a Shipmoor IC trial when you want Claim Check, Test Evidence, Code Review, Agent Skills, and the Agent Harness.

Get Shipmoor CLI

One installer. One shipmoor command. Free Community scans.

curl -fsSL https://dl.shipmoor.dev/install.sh | bash


Read the documentation

Questions about Shipmoor CLI Frequently asked questions

Clear answers about the free Scan, Shipmoor IC, local execution, and the main verification commands.

Contact sales

Our team can help with custom support, team rollouts, and self-hosted deployments. Or to get started now, explore our self-serve plans.