From superpowers
Use when about to claim work is complete and MCP cloud connection is available - calls cloud quality gates to verify superpowers workflow was followed
How this skill is triggered — by the user, by Claude, or both
Slash command
/superpowers:cloud-quality-gatesThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Before claiming completion, verify that the superpowers workflow was actually followed by running cloud-backed quality gates.
Before claiming completion, verify that the superpowers workflow was actually followed by running cloud-backed quality gates.
Core principle: Cloud gates provide objective evidence that workflow discipline was maintained.
This skill activates when bazdmeg_superpowers_gate_check MCP tool is available AND you are about to claim work is complete. If MCP tools are not available, fall back to local verification only.
Call: bazdmeg_superpowers_gate_check
sessionId: (current session)
The gate check returns 5 gates:
| Gate | What It Checks |
|---|---|
| Brainstorming Gate | Was brainstorming invoked before planning? |
| Planning Gate | Was a plan written before implementation? |
| TDD Gate | Were tests written during implementation? |
| Review Gate | Was code review requested? |
| Verification Gate | Was verification run before completion? |
| Status | Action |
|---|---|
| All GREEN | Proceed with completion |
| Any YELLOW | Note the warnings, proceed with caution, mention gaps |
| Any RED | STOP. Address the gate failure before claiming completion |
If a gate is RED:
Extends:
Pairs with:
npx claudepluginhub zerdos/superpowers --plugin superpowersCreates structured quality gates for workflow boundaries like pre-merge checklists, deployments, and phase transitions with pass/fail criteria, exact commands, and escalation procedures.
Enforces evidence-before-claims discipline: requires fresh verification (test, build, lint) before any completion claim. Prevents premature sign-offs.
Forces verification before claiming work is complete, fixed, or release-ready. Runs the exact command, reads output, and requires evidence slots (action, result, scope, risk, confidence).