Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

Qa Run · elie222 | DecimalAI

testingby elie222

Qa Run

elie222/qa-run

Source elie222/inbox-zero ↗License NOASSERTION

Run browser QA flows and write a JSON report

SkillScore77Maintained

Automatic

QA & Test Engineering pack

Last verified 8/29/2026

What it changes

Passed 18 of 23 test cases with the skill, and 14 of 23 without it.

Without skill62%

With skill77%

Checks it passed that the model alone missed

marks the blocked flow status as failed
captures at least one screenshot for the failed flow
uses the --only flag to run the specified flow identifiers

and 3 more in the full report

5 fixed15 passed either way5 still failing1 broke

Pass-rate lift: +15 pp
Tokens: -24%
Agent turns: 0%
Response time: -65%

Model gemini-3.6-flashMeasured 8/29/2026v120 of 26 cases passed with skill · 23 comparable

Tokens, turns and time describe resource use, not a dollar price.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/elie222-qa-run/SKILL.md https://app.decimal.ai/s/elie222-qa-run/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add elie222-qa-run, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/elie222-qa-run/ and pins it in .decimal/skills.lock

decimalai skills add elie222-qa-run

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/elie222-qa-run/SKILL.md https://app.decimal.ai/s/elie222-qa-run/SKILL.md

Raw file: /s/elie222-qa-run/SKILL.md ↗ · machine-readable summary at /s/elie222-qa-run.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 5 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 5

Prompt

A flow specification contains the step 'Open the Assistant settings page'. When running against base URL http://localhost:3000 with account ID 'acc-99', what exact URL should the browser navigate to?

What was checked, without and with the skill
Checked	Without skill	With skill
navigates the browser to http://localhost:3000/acc-99/automation	Missed	Passed

case-20 · gemini-3.6-flash · 8/29/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md48 linessingle file

You are the browser QA flow orchestrator. Use the flow specs in qa/browser-flows to execute tests in a real browser.

Args: $ARGUMENTS

If no args or --help is present, print usage and stop.

Usage:

/qa-run --list
/qa-run --all [--parallel] [--max-parallel=3]
/qa-run --only=flow-a,flow-b [--parallel] [--max-parallel=3]
/qa-run --group=api [--parallel]
/qa-run --all --base-url=http://localhost:3000

Base URL:

--base-url=<url> sets the Inbox Zero app URL (e.g. http://localhost:3000, https://www.getinboxzero.com, or any self-hosted URL).
If --base-url is NOT provided, ask the user which URL to test against before proceeding. Do not assume production or localhost.
When flows say "Open the Assistant settings page", navigate to <base-url>/<account-id>/automation etc.
Gmail/Outlook URLs (mail.google.com, outlook.live.com) are unaffected by this flag.

Filtering:

By default (without --all or --only), only priority: high flows run. Low-priority flows are skipped.
--all includes all flows regardless of priority.
--only=flow-a,flow-b runs exactly the specified flows regardless of priority.
--group=<name> filters to flows matching that group front matter value. Combinable with priority filtering.

Process:

Read qa/browser-flows/README.md and the selected flow files.
If --list, print each flow id + title + group + priority + resources and stop.
Determine run mode (all, only, or default high-priority). Apply --group filter if present. Fail fast if any requested ids are missing.
If --parallel, batch flows so no batch contains overlapping resources, no flow lists another in conflicts_with (missing means none), and every flow in the batch has parallel_safe: true (missing means false). If batching is not possible, run sequentially.
Execute each flow exactly as written. Use deliberate waits when moving between Gmail, Outlook, and Inbox Zero.
Record evidence. Capture at least one screenshot for every failed flow and include it in the report.
Write the JSON report to qa/browser-flows/results/<run-id>.json and save screenshots under qa/browser-flows/results/<run-id>/.
Write a companion Markdown summary to qa/browser-flows/results/<run-id>.md following the template in the README.
Print a concise summary in chat with pass/fail counts and the report path.

Output rules:

Use the JSON schema described in qa/browser-flows/README.md.

Track record

Benchmark

18 / 23

test cases passed, vs 14 / 23 without

Adoption

Users: 0
Times used: 0

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/10/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

elie222 / inbox-zero

.claude/skills/qa-run/SKILL.md

View on GitHub ↗

License NOASSERTIONImported 8/10/2026Version v1

More skills from the same source and nearby in the registry, with their headline benchmark lift.