qa-visual-criteria
Skill0 starsChanged 5 months agoMITView on GitHub
For QA (08). Gate criteria across three dimensions (functional / visual / performance) with 0-warnings acceptance, blocking/warning classification and the QA_REPORT.md structure. Use it when running the quality gate.
In this repository20 other files it gives its agents.
- docs/AGENTS.mdAGENTS.md
- adopt-groundSkill
- agent-authoringSkill
- architecture-decisionSkill
- asset-sourcingSkill
- brand-evaluation-rubricSkill
- criticSkill
- data-integritySkill
- diagrams-gateSkill
- drift-protocolSkill
- env-reproducibleSkill
- orchestrator-bridgeSkill
- project-cleanupSkill
- qa-eyesSkill
- research-methodologySkill
- security-auditSkill
- setupSkill
- smoke-testSkill
- technical-documentationSkill
- visualizerSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 13
If it iterates more than 3 times on the same problem → stop and escalate.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
SKILL: qa-visual-criteria
For: Agent 08 — QA
Version: 1.0.0
Core principle
Acceptance gate: 0 warnings — not "no blocking", but nothing at all. Strategy: blocking first in a loop → then warnings down to 0. If it iterates more than 3 times on the same problem → stop and escalate.
Classification
BLOCKING: product doesn't work in the main flow, data loss, a critical component doesn't render, horizontal overflow, error with no message, performance >3s with no feedback, security compromised. WARNING: minor visual inconsistency, no loading skeleton, truncated text, spacing outside tokens, typography off-scale, non-critical degraded performance, basic accessibility missing.
Visual dimension
Typography: fonts and sizes per tokens, consistent weight per level. Layout: no overflow at any breakpoint, consistent spacing, no overlapping elements. Components: colors per tokens, consistent radius, states present (default/hover/active/disabled).
Functional dimension
States always present: loading, error, empty, success. Main flow end-to-end with no errors or dead ends. Forms: visible validation, destructive actions with confirmation, no duplicates.
Performance dimension — only measure, never propose solutions
Initial load: FCP <2s. Navigation: transitions <500ms. APIs: <500ms for simple queries, blocking if >3s with no feedback. Report with exact numbers and measurement conditions. Tools: DevTools Network/Performance, Playwright, Console.time().
QA_REPORT.md structure
Summary (X blocking, Y warnings, status) → Blocking findings (ID, description, evidence, impact, proposal, status) → Warnings (same structure) → Performance metrics → Iterations → Final status
Mistakes to avoid
× Passed with pending warnings × Proposing performance solutions without measuring × Continuing the loop after 3 failed attempts × Reporting "slow" without an exact number × Running destructive actions during QA
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.