test-suite
Skill2 starsChanged 9 months agoApache-2.0View on GitHub
Multi-phase UI testing orchestration for SoNash. Runs smoke tests, feature protocol tests, security checks, and performance checks against any deployment URL using Playwright MCP or Chrome Extension. Navigate to target URL and verify app loads on key routes. 1. mcp__playwright__browser_navigate to target URL 2. Verify page loads (check for app shell elements) 3. Navigate to each route: /, /notebook, /admin, /journal, /meetings/all 4. On each page: - mcp__playwright__browser_snapshot — check for page content - mcp__playwright__browser_console_messages — check for errors -…
3 other projects use a version of this.
Adopted by
Most recently taken up by
- mgiovani/cc-arsenaltheir own version, first seen Sep 25, 2026
- poorvith-mp/skills-developertheir own version, first seen Aug 19, 2026
- Kastalien-Research/thoughtboxtheir own version, first seen Jul 3, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- Kastalien-Research/thoughtbox★ 64same name
- mgiovani/cc-arsenal★ 8same name
- poorvith-mp/skills-developer★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository81 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- llms.txtllms.txt
- find-skillsSkill
- add-debtSkill
- alertsSkill
- analyzeSkill
- artifacts-builderSkill
- audit-agent-qualitySkill
- audit-aggregatorSkill
- audit-ai-optimizationSkill
- audit-codeSkill
- audit-comprehensiveSkill
- audit-documentationSkill
- audit-engineering-productivitySkill
- audit-enhancementsSkill
- audit-healthSkill
- audit-performanceSkill
- audit-processSkill
- audit-refactoringSkill
- audit-securitySkill
- brainstormSkill
- checkpointSkill
- code-reviewerSkill
- comprehensive-ecosystem-auditSkill
- content-research-writerSkill
- convergence-loopSkill
- create-auditSkill
- data-effectiveness-auditSkill
- debt-runnerSkill
- decrypt-secretsSkill
- deep-planSkill
- deep-researchSkill
- developer-growth-analysisSkill
- doc-ecosystem-auditSkill
- doc-optimizerSkill
- docs-maintainSkill
- document-analysisSkill
- ecosystem-healthSkill
- Excel AnalysisSkill
- find-skillsSkill
- frontend-designSkill
- gh-fix-ciSkill
- github-healthSkill
- health-ecosystem-auditSkill
- hook-ecosystem-auditSkill
- market-research-reportsSkill
- mcp-builderSkill
- media-analysisSkill
- multi-ai-auditSkill
- pre-commit-fixerSkill
- pr-ecosystem-auditSkill
- pr-retroSkill
- pr-reviewSkill
- quick-fixSkill
- recallSkill
- repo-analysisSkill
- repo-synthesisSkill
- script-ecosystem-auditSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
11 rules. 0 are common, 8 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
8
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 23
When the task doesn't match this skill's scope -- check related skills
- Distinctive: line 24
When a more specialized skill exists for the specific task
- Distinctive: line 78
Verify page loads (check for app shell elements)
- Distinctive: line 86
GATE: If ANY page fails to load → ABORT entire test run
- Distinctive: line 157
Check no email addresses or raw user IDs in request URLs/bodies
- Very specific: line 158
Check security headers (via mcp__playwright__browser_network_requests):
- Very specific: line 162
If headers unavailable in tool output → record as skipped (reason: "headers_unavailable_in_tool")
- Distinctive: line 164
Check console for exposed secrets:
- Distinctive: line 165
No API keys, tokens, or credentials in console output
- Distinctive: line 180
Check total page weight:
- Very specific: line 319
Always install Playwright first: npx playwright install (if browser_navigate fails)
“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
/test-suite — Unified Testing Suite
<!-- prettier-ignore-start -->Document Version: 1.0 Last Updated: 2026-02-07 Status: ACTIVE
<!-- prettier-ignore-end -->When to Use
- Tasks related to test-suite
- User explicitly invokes
/test-suite
When NOT to Use
- When the task doesn't match this skill's scope -- check related skills
- When a more specialized skill exists for the specific task
Purpose
Multi-phase UI testing orchestration for SoNash. Runs smoke tests, feature protocol tests, security checks, and performance checks against any deployment URL using Playwright MCP or Chrome Extension.
AI Instructions
Invocation
/test-suite [scope] [--url=<target>] [--protocol=<name>] [--chrome]Scopes
| Flag | Description | Duration |
--smoke | Quick health check (all pages load) | ~1 min |
--feature | Run all feature protocols | ~5-10 min |
--security | Security-focused checks | ~2 min |
--performance | Performance checks | ~2 min |
--full | All phases | ~10-15 min |
--protocol=NAME | Run single protocol by name | ~30s |
--post-deploy | Production smoke after merge | ~1 min |
URL Resolution (in order)
--url=URLflag (explicit)- PR preview URL via
gh pr view --json url(auto-detect) https://sonash-app.web.app(production fallback)
Browser Selection
| Condition | Tool |
| Default | Playwright MCP (mcp__playwright__*) |
Step has "auth_required": true | Chrome Extension |
Step has "tool": "chrome" | Chrome Extension |
User passes --chrome | Chrome Extension for all steps |
Execution Pipeline
Phase 1: SMOKE
Navigate to target URL and verify app loads on key routes.
Steps:
mcp__playwright__browser_navigateto target URL- Verify page loads (check for app shell elements)
- Navigate to each route:
/,/notebook,/admin,/journal,/meetings/all - On each page:
mcp__playwright__browser_snapshot— check for page contentmcp__playwright__browser_console_messages— check for errorsmcp__playwright__browser_take_screenshot— save to.claude/test-results/screenshots/
- GATE: If ANY page fails to load → ABORT entire test run
Record each step as JSONL:
{
"test_id": "SMOKE-01",
"protocol": "smoke",
"phase": "smoke",
"name": "Homepage loads",
"status": "pass",
"duration_ms": 850,
"assertions": [{ "type": "text_visible", "value": "SoNash", "passed": true }],
"screenshot": ".claude/test-results/screenshots/SMOKE-01.png",
"error_message": null,
"timestamp": "2026-02-07T12:00:00Z",
"target_url": "https://sonash-app.web.app"
}Phase 2: FEATURE PROTOCOLS
Execute feature test protocols from .claude/test-protocols/.
Steps:
- Scan
.claude/test-protocols/*.protocol.json(exclude_base.protocol.json) - Sort by priority:
core→admin→secondary→edge - For
--smokescope: skip this phase - For
--featurescope: run all protocols matching priority filter - For
--protocol=NAMEscope: run only the named protocol - For each protocol file: a. Read the JSON file b. Check prerequisites (auth
level, skip_on_preview) c. Execute each step sequentially:
- For
navigateaction: resolvestep.valueagainst the target URL origin (if it starts with/, build${new URL(target_url).origin}${step.value}) then callmcp__playwright__browser_navigate - For
clickaction:mcp__playwright__browser_clickwith selector - For
typeaction:mcp__playwright__browser_typewith selector + value - For
selectaction:mcp__playwright__browser_select_option - For
snapshotaction:mcp__playwright__browser_snapshot - For
evaluateaction:mcp__playwright__browser_evaluated. After each step, check assertions: text_visible: snapshot contains expected texttext_not_visible: snapshot does NOT contain textelement_exists: snapshot contains matching selectorelement_count: evaluatedocument.querySelectorAll(selector).lengthno_console_errors:mcp__playwright__browser_console_messageshas no errorsnetwork_no_pii:mcp__playwright__browser_network_requestshas no emails/userIdsurl_contains: current URL includes stringurl_redirected: navigation redirected to expected URLpage_load_under: evaluate page load duration via Navigation Timing API:(() => { const nav = performance.getEntriesByType('navigation')[0]; return nav ? nav.duration : null; })()and assert result is non-null and < threshold (ms)evaluate_truthy: evaluate the JavaScript expression and assert it returns truthy e. Ifscreenshot: true, take screenshot f. If step hasbrowser_resizeproperty ({ "width": N, "height": N }), callmcp__playwright__browser_resizebefore executing the step g. Record result as JSONL line
- For
- Continue on failure — collect all results, do not abort
Phase 3: SECURITY
Security-focused checks.
Steps:
- Navigate to
/adminwithout auth → verify redirect to login - Scan network requests for PII:
mcp__playwright__browser_network_requests- Check no email addresses or raw user IDs in request URLs/bodies
- Check security headers (via
mcp__playwright__browser_network_requests):- Locate the document response for the target route
- Assert headers include:
Content-Security-Policy,X-Frame-Options(orframe-ancestorsvia CSP),Strict-Transport-Security - If headers unavailable in tool output → record as skipped
(
reason: "headers_unavailable_in_tool")
- Check console for exposed secrets:
- No API keys, tokens, or credentials in console output
Phase 4: PERFORMANCE
Performance checks.
Steps:
- Navigate to homepage, measure load time:
mcp__playwright__browser_evaluate:(() => { const nav = performance.getEntriesByType('navigation')[0]; return nav ? nav.duration : null; })()- Threshold: < 3000ms (skip if
null)
- Count network requests:
mcp__playwright__browser_network_requests- Flag if > 50 requests
- Check total page weight:
- Sum transfer sizes from network requests
- Flag if > 5MB
- Take screenshot for visual baseline comparison
Phase 5: REPORT
Generate results files.
Steps:
- Collect all JSONL lines from phases 1-4
- Write to
.claude/test-results/{YYYY-MM-DD}-{scope}.jsonl - Generate markdown report at
.claude/test-results/{YYYY-MM-DD}-{scope}-report.md:
# Test Suite Report — {date} ({scope} scope)
**Target:** {url} **Duration:** {total_duration} **Result:**
{pass_count}/{total} passed, {fail_count} failed, {skip_count} skipped
## Summary
| Phase | Pass | Fail | Skip | Total |
| ----------- | ---- | ---- | ---- | ----- |
| Smoke | ... | ... | ... | ... |
| Feature | ... | ... | ... | ... |
| Security | ... | ... | ... | ... |
| Performance | ... | ... | ... | ... |
## Failures
### {test_id}: {name}
- **Protocol:** {protocol}
- **Assertion:** {failed_assertion}
- **Screenshot:** {screenshot_path}
- **Suggested Fix:** {suggestion}
## Skipped Tests
| Test | Reason |
| ---- | --------------------------------------------- |
| ... | manual_only / skip_on_preview / auth_required |
## All Passed Tests
<details><summary>Click to expand ({count} tests)</summary>
| Test ID | Protocol | Name | Duration |
| ------- | -------- | ---- | -------- |
| ... | ... | ... | ... |
</details>- Print summary to conversation:
- Total pass/fail/skip counts
- List of failures with one-line descriptions
- Path to full report
Protocol File Format
Protocol files live at .claude/test-protocols/{feature-name}.protocol.json.
{
"protocol_version": "1.0",
"feature": "Feature Name",
"created": "2026-02-07",
"created_session": 141,
"priority": "core|admin|secondary|edge",
"target_route": "/route",
"prerequisites": {
"auth": "none|user|admin",
"description": "Human-readable prereq"
},
"steps": [
{
"id": "FEAT-01",
"name": "Step description",
"action": "navigate|click|type|select|snapshot|evaluate",
"selector": "text=Button or [data-testid='x']",
"value": "text to type (for type action)",
"wait_for": "text=Expected or network_idle",
"screenshot": true,
"auth_required": false,
"skip_on_preview": false,
"assertions": [
{ "type": "text_visible", "value": "Expected text" },
{ "type": "element_exists", "selector": "[data-testid='x']" },
{ "type": "no_console_errors" }
]
}
],
"error_scenarios": [
{
"id": "FEAT-E1",
"name": "Edge case name",
"description": "What to verify",
"manual_only": true,
"expected": "Expected behavior"
}
]
}Assertion Types Reference
| Type | Description | MCP Tool |
text_visible | Text appears on page | browser_snapshot |
text_not_visible | Text does NOT appear | browser_snapshot |
element_exists | CSS selector matches | browser_snapshot |
element_count | N elements match selector | browser_evaluate |
no_console_errors | No error-level console messages | browser_console_messages |
network_no_pii | No PII in network requests | browser_network_requests |
url_contains | URL contains string | browser_snapshot |
url_redirected | Redirected to expected URL | browser_navigate |
screenshot_baseline | Visual comparison | browser_take_screenshot |
response_status | Network response code | browser_network_requests |
page_load_under | Load time less than N ms | browser_evaluate |
Results Location
- JSONL:
.claude/test-results/{YYYY-MM-DD}-{scope}.jsonl - Report:
.claude/test-results/{YYYY-MM-DD}-{scope}-report.md - Screenshots:
.claude/test-results/screenshots/{test_id}.png
Notes
- Preview channels deploy hosting only — protocols with
skip_on_preview: truetest backend integration and should be skipped on preview URLs - Steps with
auth_required: trueneed Chrome Extension (real login state) or should be skipped - The
_base.protocol.jsondefines common assertions applied to all protocols - Always install Playwright first:
npx playwright install(if browser_navigate fails)
Version History
| Version | Date | Description |
| 1.0 | 2026-02-25 | Initial implementation |
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
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.

