mirror of https://github.com/Donchitos/Claude-Code-Game-Studios.git synced 2026-10-16 01:37:05 +00:00

Files

T

DonchitosandClaude Opus 5 d05699707f Release 1.1.0 — one config file, and a process weight you choose

`project.yaml` at the repo root is now the single source of truth for a
project's configuration, replacing `production/stage.txt`,
`production/review-mode.txt` and `.claude/docs/technical-preferences.md` (all
three still read as a fallback; nothing is force-deleted).

**`modes.rigor` — one question instead of six.** `minimal`, `standard` or
`full` fronts six knobs: `modes.workflow`, `docs.density`, `qa.level`,
`modes.story_granularity`, `modes.review_mode` and `team.size`. It defaults to
`minimal`, which is a change from 1.0's behaviour: the heavier tier was
measured costing several times more to reach working code without producing a
better result, and returning nothing when a fresh developer inherited the
project. Pin `modes.rigor: standard` to keep 1.0's process level.

Also in this release:

- `/settings` to view or change any setting, and `project.local.yaml` for
  per-developer preferences that should not be committed.
- `system_overrides` to hold one system to a higher standard than the rest of
  the project, and `testing.strict` to set per test type whether missing
  evidence blocks a story or warns.
- A story that changes anything the player sees is not closed until the game
  has been launched, observed, and a screenshot retained in
  `production/qa/evidence/`. A parse check is not a run. This holds at every
  rigor level, including `minimal`, where automated tests are waived but the
  look is not.
- The code root is engine-specific and enforced as such: Godot `src/`, Unity
  `Assets/`, Unreal `Source/<Module>/`. Two of the three will not build a
  project laid out any other way.
- Migration tooling for existing projects: `.claude/scripts/migrate-v1-config.sh`
  converts a 1.0 layout, non-destructively by default, with a report to review
  before `--finalize` deletes anything.
- Ask-before-write holds regardless of personal automation settings.

Upgrading from 1.0 or earlier: see UPGRADING.md. Full list of changes:
CHANGELOG.md.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016xvzG3i7pWQg4LjyDwWMXd

2026-09-23 12:25:11 +10:00

7.8 KiB

Raw Permalink Blame History

CCGS Skill Testing Framework

Quality assurance infrastructure for the Claude Code Game Studios framework. Tests the skills and agents themselves — not any game built with them.

This ships as part of the template, and it is meant to. CCGS is a template you are expected to customize — edit a skill, add your own, retune an agent. This folder is how you check that what you changed still holds up: catalog.yaml tracks all 74 skills and 49 agents, quality-rubric.md defines per-category pass/fail metrics, and templates/ gives you the spec format for anything new you write. Driven by /skill-test and /skill-improve.

If you remove it, two of the four /skill-test modes stop working.

Mode Without this folder

/skill-test static Works — the 7 structural checks read SKILL.md only

/skill-test audit Degrades — reports that no catalog exists

/skill-test spec Breaks — the behavioral specs live here

/skill-test category Breaks — reads quality-rubric.md from here

/skill-improve Degrades — its test-fix-retest loop loses the spec pass

Earlier revisions of this file called the folder "optional" and said "nothing in .claude/ depends on it." That was wrong — /skill-test references this directory 16 times and /skill-improve once. Removing it is a real trade, not a free cleanup, so the table above states the actual cost.

Mode	Without this folder
`/skill-test static`	Works — the 7 structural checks read `SKILL.md` only
`/skill-test audit`	Degrades — reports that no catalog exists
`/skill-test spec`	Breaks — the behavioral specs live here
`/skill-test category`	Breaks — reads `quality-rubric.md` from here
`/skill-improve`	Degrades — its test-fix-retest loop loses the spec pass

What's in here

CCGS Skill Testing Framework/
├── README.md              ← you are here
├── CLAUDE.md              ← tells Claude how to use this framework
├── catalog.yaml           ← master registry: all 74 skills + 49 agents, coverage tracking
├── quality-rubric.md      ← category-specific pass/fail metrics for /skill-test category
│
├── skills/                ← behavioral spec files for skills (one per skill)
│   ├── gate/              ← gate category specs
│   ├── review/            ← review category specs
│   ├── authoring/         ← authoring category specs
│   ├── readiness/         ← readiness category specs
│   ├── pipeline/          ← pipeline category specs
│   ├── analysis/          ← analysis category specs
│   ├── team/              ← team category specs
│   ├── sprint/            ← sprint category specs
│   └── utility/           ← utility category specs
│
├── agents/                ← behavioral spec files for agents (one per agent)
│   ├── directors/         ← creative-director, technical-director, producer, art-director
│   ├── leads/             ← lead-programmer, narrative-director, audio-director, game-designer, systems-designer, level-designer, qa-lead
│   ├── specialists/       ← gameplay/engine/network/AI/tools/UI programmers, technical-artist, sound-designer, ux-designer, performance-analyst, prototyper, writer, world-builder
│   ├── engine/            ← engine-specific specialists, split per engine:
│   │   ├── godot/         ← Godot-specific specialists
│   │   ├── unity/         ← Unity-specific specialists
│   │   └── unreal/        ← Unreal-specific specialists
│   ├── operations/        ← devops, release, live-ops, community, analytics, economy, localization
│   └── qa/                ← qa-tester, security-engineer, accessibility-specialist
│
├── templates/             ← spec file templates for writing new specs
│   ├── skill-test-spec.md ← template for skill behavioral specs
│   └── agent-test-spec.md ← template for agent behavioral specs
│
└── results/               ← test run outputs (written by /skill-test spec, gitignored)

How to use it

All testing is driven by two skills already in the framework:

Check structural compliance

/skill-test static [skill-name]     # Check one skill (7 checks)
/skill-test static all              # Check all 74 skills

Run a behavioral spec test

/skill-test spec gate-check         # Evaluate a skill against its written spec
/skill-test spec design-review

Check against category rubric

/skill-test category gate-check     # Evaluate one skill against its category metrics
/skill-test category all            # Run rubric checks across all categorized skills

See full coverage picture

/skill-test audit                   # Skills + agents: has-spec, last tested, result

Improve a failing skill

/skill-improve gate-check           # Test → diagnose → propose fix → retest loop

Skill categories

Category	Skills	Key metrics
`gate`	gate-check	Review mode read, full/lean/solo director panel, no auto-advance
`review`	design-review, architecture-review, review-all-gdds	Read-only, 8-section check, correct verdicts
`authoring`	design-system, quick-design, art-bible, create-architecture, …	Section-by-section May-I-write, skeleton-first
`readiness`	story-readiness, story-done	Blockers surfaced, director gate in full mode
`pipeline`	create-epics, create-stories, dev-story, map-systems, …	Upstream dependency check, handoff path clear
`analysis`	consistency-check, balance-check, code-review, tech-debt, …	Read-only report, verdict keyword, no writes
`team`	team-combat, team-narrative, team-audio, …	All required agents spawned, blocked surfaced
`sprint`	sprint-plan, sprint-status, milestone-review, …	Reads sprint data, status keywords present
`utility`	start, adopt, hotfix, localize, setup-engine, …	Passes static checks

Agent tiers

Tier (directory)	Agents
`directors`	creative-director, technical-director, producer, art-director
`leads`	lead-programmer, narrative-director, audio-director, game-designer, systems-designer, level-designer, qa-lead
`specialists`	gameplay-programmer, engine-programmer, network-programmer, ai-programmer, tools-programmer, ui-programmer, ux-designer, technical-artist, sound-designer, performance-analyst, prototyper, writer, world-builder
`engine/godot`	godot-specialist, godot-gdscript-specialist, godot-csharp-specialist, godot-shader-specialist, godot-gdextension-specialist
`engine/unity`	unity-specialist, unity-ui-specialist, unity-shader-specialist, unity-dots-specialist, unity-addressables-specialist
`engine/unreal`	unreal-specialist, ue-gas-specialist, ue-replication-specialist, ue-umg-specialist, ue-blueprint-specialist
`operations`	devops-engineer, release-manager, live-ops-designer, community-manager, analytics-engineer, economy-designer, localization-lead
`qa`	qa-tester, security-engineer, accessibility-specialist

Updating the catalog

catalog.yaml tracks test coverage for every skill and agent. After running a test:

/skill-test spec [name] will offer to update last_spec and last_spec_result
/skill-test category [name] will offer to update last_category and last_category_result
last_static and last_static_result are updated manually or via /skill-improve

Writing a new spec

Find the spec template at templates/skill-test-spec.md
Copy it to skills/[category]/[skill-name].md
Update the spec: field in catalog.yaml to point to the new file
Run /skill-test spec [skill-name] to validate it

Removing this framework

This folder has no hooks into the main project. To remove:

rm -rf "CCGS Skill Testing Framework"

The skills /skill-test and /skill-improve will still function — they'll simply report that catalog.yaml is missing and suggest running /skill-test audit to initialize it.

7.8 KiB Raw Permalink Blame History