full-qa

An autonomous quality-assurance process for web and other projects that follows a test plan, runs tests in a browser, fixes found bugs, and creates a final report. Quality assurance means checking that software behaves as required.

In plain words
What is it for?
Use it before a release or after major changes to execute a written test plan, test web behavior through Chrome, repair defects, repeat failed checks, and report the outcome.
Why use it?
It combines testing, bug fixing, and repeated checks so failures are addressed during the same review. It also limits autonomous browser work to test environments when real logged-in accounts are involved.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/devotts/build-it/full-qa
Any agent
npx skills add DevOtts/build-it --skill full-qa
Clone the repo
git clone --depth 1 https://github.com/DevOtts/build-it

Made for: Claude Code, Codex.

Per session 82 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,444 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 2 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00082 $0.03444
Opus 5 $0.00041 $0.01722
Sonnet 5 $0.00016 $0.00689
Haiku 4.5 $0.00008 $0.00344

Measured 2d ago against content hash ca58badff4f9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

full-qa scanned grade C with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Downloads and executes remote codehighSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

curl -s "$CDP_URL/json/version" | python3 -c "import sys,json; d=json.load(sys.stdin); print('Chrome:', d.get('Browser','?'))"

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -s -o /dev/null -w "%{http_code}" <SERVICE_URL>/health
plugins/build-it/skills/full-qa/SKILL.md · 320 lines

How it starts

The opening of the file, as written. The whole thing — 320 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/full-qa — Autonomous Full QA Suite

You are running an autonomous, project-agnostic QA pipeline. You read a test plan, execute every test, fix bugs as you find them, and deliver a final pass/fail report — without stopping between steps unless a destructive or truly ambiguous action requires confirmation.

This skill incorporates Chrome CDP browser control and iterative bug-fix cycles natively. You do not need to invoke /chrome-cdp-control or /iterate separately.

CDP mechanics live in one shared reference — read ../references/cdp-core.md (relative to this skill's base directory) before any browser work. It canonically owns endpoint resolution (CDP_URL/APP_URL env → grounding/test plan → defaults), the action template, tab selection, the selector ladder, waits, and the failure protocol. Never restate them here; never hardcode a CDP endpoint or app port — parallel runs collide on hardcoded values.

Route guard (cdp-core.md §9): autonomous mode is for TEST ENVIRONMENTS ONLY. If any test case targets the user's authenticated real-Chrome session (their logged-in accounts — posting, sending, buying), REFUSE to run that case autonomously and re-route it to /chrome-cdp-control, whose per-write gate requires explicit user confirmation for each write. Unattended (user asleep) that case ends as BLOCKED/deferred with the reason. There is no autonomous write path on an authenticated session.


Step 0 — INGEST THE TEST PLAN

First action: identify the test plan.

  • If the user passed a file path (e.g., /full-qa ./E2E-test-plan.md), read that file immediately.
  • If the user pasted tests inline, extract them from the conversation.
  • If neither: ask the user ONE question: "Where is the test plan? Paste it or give me a file path."

From the test plan, extract:

  1. Service URLs — all localhost:PORT or external URLs mentioned
  2. Auth credentials — any test user emails/passwords
  3. Test cases — each scenario with its ID, steps, and pass criteria
  4. Setup steps — any DB resets, seed commands, or imports required before tests
  5. Stack info — language, framework, DB type (for tailoring fix strategies)

Read the full file on GitHub · 320 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 320 lines · 82 tokens per session scan C ca58badff4f9

Subscribe to this mod's changes

full-qa is a skill published in the GitHub repository DevOtts/build-it (14 stars, last pushed 1mo ago), licensed MIT. It adds 82 tokens to every session and 3,444 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.