Does it handle both technical QA and playtesting?

Yes, as separate tracks. Technical checks examine behavior against accepted rules; playtest protocols and findings examine actual player actions, comprehension and experience.

What should I provide for a regression test?

The accepted behavior, candidate source or build, relevant engine/platform context and available test tools. Include existing results or reproduction details when available.

Can it perform a human playtest for me?

It can prepare a concrete protocol and analyze observations you provide. Actual sessions require participants and suitable recording or observation arrangements; it does not invent player feedback.

Do I need the other Game Squad specialists?

No. The complete independent QA/playtest method and reference are included. Its underlying method is also part of Game Squad Agent Skills; companion roles are optional.

What does the preview show?

An illustrative checkpoint test plan with independently derived expected states. It shows the planning format, rather than a recorded model session or executed game tests.

Can I use the test plans commercially?

Yes. Personal and commercial project use is permitted. Keep purchased source materials private under Agensi's personal, non-transferable source license.

Game QA & Playtest Specialist

Example session with this skill installed

Design a QA evidence pack, without running anything. Offline rule: collecting a stable pickup ID awards exactly one point once. Checkpoint restores both score and collected IDs from the saved moment. Test sequence: collect A, save, collect B, restore, collect A, collect B, restore again. Missing or invalid saves must leave the live state unchanged. Source, build identity, engine and hardware are unavailable. Provide independent expected states, negative cases, binding requirements and next action. Keep proposed tests separate from observed results.

Read your context and instructions
Compiled the game qa

Illustrative preview: this representative editorial result is not a recorded runtime session.

task_id: PICKUP-QA-1
artifact_id: CHECKPOINT-EVIDENCE-PLAN
version: 1
status: NEEDS_EVIDENCE
producer / recipient: game QA / fixture and integration owner
objective / scope: Plan functional evidence for stable pickup rewards and checkpoint atomicity.
inputs_and_sources: The supplied pickup/checkpoint rules; no source, candidate or raw result inspected.
confirmed_decisions: One point per fresh ID; restore uses saved score and IDs; invalid/missing save cannot mutate live state.
recommendations_and_pending: Acquire immutable source/scene/settings/data and actual build identities, engine/environment, spec and external harness binding.
criteria: PICK-1 uniqueness, SAVE-1 checkpoint restoration, SAVE-2 invalid-save atomicity
next_action: Owner supplies candidate and environment; collector executes the planned checks without repairing the frozen candidate.

Independent rule oracle, not execution
Start: score 0, collected {}.

Collect A -> score 1, {A}.
Save S -> S stores score 1, {A}.
Collect B -> score 2, {A,B}.
Restore S -> score 1, {A}.
Collect A -> rejected reward; score 1, {A}.
Collect B -> award once; score 2, {A,B}.
Restore S again -> score 1, {A}.
These expected states follow the accepted rules, not an implementation log.

Planned coverage

Q-01 / PICK-1: fresh A, repeated A and distinct B; record returned result, score and ID set after each action.
Q-02 / SAVE-1: execute the full sequence above, including repeated restore; compare every expected state.
Q-03 / SAVE-2: missing file, truncated payload, wrong field type and unknown schema against the accepted validation policy. Capture pre-state and verify no live mutation; keep any test payload outside the candidate. Unknown schema migration behavior needs a decision if the task requires migration.
Q-04 / integration: trigger the actual scene pickup path, including repeated collision/event delivery. A pure rule fixture cannot establish collision hookup or save-file I/O.
Q-05 / exported build: repeat the relevant journey in the requested device/build if that acceptance scope is later supplied. No platform requirement is inferred.

For each check record: check/criterion ID, source/build/asset/data identity, specification and harness identity, input/preconditions, independent expected observation, exact steps, actual environment/version, time, exit status, sanitized raw observations and PASS/FAIL/PENDING. Verify the frozen candidate before and after collection; failures belong in the pack, never silently patched.

Current result table: Q-01 through Q-05 PENDING. No engine, filesystem, save fault injection or hardware test ran. No defect is asserted without an actual observation. This pack plans functional QA only; it provides no player-experience or fun conclusion.

Connects securely to your tools. The creator never sees your data.

About this skill

Know which game problem you are solving next.

Game QA & Playtest Specialist helps you separate a functional defect from an experience question. Use it to challenge a gameplay change, plan regression coverage, reproduce a save or input issue, or turn real player observations into a useful design finding.

The method derives expected behavior from your accepted rules, then chooses checks that can expose the relevant failure. It tracks the build and conditions each result belongs to. Player research follows a separate path, with a clear question, neutral tasks and a distinction between what participants did, what they said and your interpretation.

Ask for

Targeted rule, state, input, save and integration cases with expected results.
Reproducible defect records and a clear coverage summary.
A playtest protocol or synthesis of supplied observations that informs the next design decision.

For example, test whether collecting an item twice or restoring a checkpoint can award a duplicate point. Separately, ask whether players understand what the checkpoint preserves. The specialist keeps those questions connected without treating a green test as proof of player comprehension.

Provide the accepted rules, identified source/build and tools, or the player observations you want analyzed. The complete standalone method and reference are included. You supply AI access. Executing game tests needs the project environment; conducting playtests needs real participants. Without them, the specialist delivers the filled test or study plan you can run next.

How to use

Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.

Reviews

No reviews yet

Be one of the first to try it. Every listed skill passes our trust checks below.

Security scanned

Passed our 8-point scan before listing

Fresh listing

Recently published to Agensi

Money-back guarantee

Doesn't do what its page promises? You get your money back.

Game QA & Playtest Specialist