Operations and API

Use one target throughout an investigation: project, environment, app or service, and deployment. Recheck it before a mutation, especially in production.

Task	Surface	CLI coverage
Deploy source and inspect readiness	App deployment page or CLI	push, readiness, slide, deployments
Read deployment output	Deployment detail or CLI	logs --deployment ID
Read running app logs	App logs	CLI 0.0.4/server 0.0.120: logs, --follow
Inspect Sails models and helpers	Helm	CLI 0.0.4/server 0.0.120 run executes commands, not model-console JavaScript
Query or migrate a database	Dock	No dedicated Dock command
Inspect and operate background jobs	Quest	No dedicated Quest command
Restore a database backup	Service backups	backup:restore ID --writes-paused
Roll back an app	Deployment history	No rollback command

Authorization and review

Browser actions and authenticated API requests apply server-side team and target checks. Deploy-token management has a separate availability boundary; do not assume those tokens authenticate deployment or operational requests. Never transfer a credential into a URL, chat transcript, screenshot, or committed script.

For a database change, inspect the current schema and take a recoverable backup first. For a production command, use Helm's write-arm rules. For a job, review its declared inputs and side effects. Permission to open a tool does not make its work read-only.

Wait and verify

Distinguish request acceptance, progress, and terminal outcome. Retain an operation or deployment ID when one is returned. After a timeout or disconnect, inspect that operation and its actual effect before repeating the request.

For database restore, pause app and external writers before invoking:

slipway backup:restore BACKUP_ID --writes-paused

The CLI polls the returned restore operation for up to an hour. If it remains pending, inspect /api/v1/restore-operations/OPERATION_ID with an authorized client and keep writers paused. A completed operation still requires verification of the restored database before traffic resumes. See database backups.

API references

Use the tool-specific contract rather than guessing a URL from a CLI command:

• Dock API for query, schema, migration, import, and export requests.
• Quest for job compatibility and available operational controls.
• Deploy tokens for scoped pipeline deployment.
• CLI workflow for structured logs/run output, exit codes, and unconfirmed outcomes.

These transports have different capabilities. The CLI 0.0.4 adds app:inspect, exact-target-approved app:restart, private command receipts, run:arm, run:cancel, and metadata-only run:history through existing server APIs. See the CLI workflow for their release compatibility boundary, completion evidence, and cancellation limits. It has no general operation lookup, automatic replay, or resume interface.

Checks, recovery and analytics

• Slipway CI: local source checks and server deployment verification.
• Restore tests: test managed PostgreSQL backups in a temporary database.
• Wake: product analytics, consent, goals and revenue.
• Instance maintenance: inspect maintenance jobs and recover control-plane storage.