The problem
Your most strategic engineers are copying and pasting scripts from Slack.
As teams grow, production access concentrates in a few senior people. Every database migration, every config change, every hotfix, routed through the same engineers. They stop building and start executing other people's scripts. It starts as a few requests a day. It ends as a full-time job.
~24 min per request. 15+ requests/day. One engineer gone.
HOW IT WORKS
Submit. Route. Review. Execute.
Hoop replaces the Slack back-and-forth with a structured approval flow. The person who needs the command executed submits it directly. Hoop routes it to the right approver with full context. No copy-pasting. No ambiguity about which environment. No lost scripts in thread replies.
Through the Hoop web terminal, CLI, API, or Jira ticket. The command is held. Nothing executes yet.
Hoop routes the request to the correct approver based on resource, team, and escalation rules. Slack, Teams, or a Jira ticket with the right assignee.
The approver sees who is requesting, which resource, and the exact command. An LLM risk analysis flags schema impact, table sizes, and potential issues before the human even looks.
Approve for immediate execution, deny with feedback, or approve with a time window. "Run this at 2 AM during the maintenance window." The user gets a real-time notification either way.
INTEGRATIONS
Approval happens where your team already works.
Hoop fits into your existing workflow, whether that's Slack threads, Teams channels, or Jira tickets with established change management processes. No new tool for approvers to learn. No new dashboard to check.
Approval requests land in the right channel with full context. Approvers click a button. No context switching, no login required. Real-time notification back to the requester.
For teams with established change management workflows. Hoop runs headless: tickets are created in Jira with the script, routed through your existing approval process, and executed automatically on approval.
The Hoop web app shows every pending request with full session context. The API lets you build custom approval workflows or integrate with internal tools.
AI RISK ANALYSIS
The approver sees the risk before reading the script.
Hoop attaches an AI-generated risk analysis to every approval request. The LLM has access to infrastructure context like schema, table sizes, and active connections, flagging potential issues before the human reviewer even opens the request. A DELETE on a 200M-row table gets flagged differently than a DELETE on a 50-row config table.
WHEN TO USE WHAT
Command-level control or time-window access. Choose per resource.
Command Approval and Just-in-Time Access solve different problems. Use Command Approval for write operations and sensitive queries. Use JIT for debugging sessions and read-only access. Most teams use both.
Command Approval
Scope
Each command reviewed individually
Best for
Write operations, migrations, hotfixes
Security
Highest. Nothing runs without explicit approval
Overhead
Higher, but AI risk analysis reduces reviewer load
Just-in-Time Access
Scope
Time window. All commands allowed during the session
Best for
Debugging sessions, read-only access, on-call
Security
High. Access is temporary, not permanent
Overhead
Lower. One approval per session, not per command
ORGANIZATIONAL IMPACT
From manual script execution to automated approval workflows.
Every approval, every denial, every scheduled execution becomes audit evidence automatically. Your SRE team gets their time back. Your compliance team gets the trail they need.
How many hours does your team spend executing other people's scripts?
Most teams don't track it until they realize a senior engineer is spending half their week on it. We can show you the impact in your first week.