
Control Manifest Builder
Turns accepted architecture decisions into a flat must-do and never-do rules sheet for programmers
- Repository
- Donchitos/Claude-Code-Game-Studios
- Stars
- 32,965
- Licence
- MIT
- Files
- 1 file
What it does
Once installed, the agent can read a game project's accepted architecture decision records (ADRs), the project's technical preferences, and engine reference docs, then write a short rules sheet called a control manifest. The sheet lists what programmers must do and must never do, grouped by architectural layer (foundation, core, feature, presentation), with each rule traced to its source. It asks for your approval before writing the file docs/architecture/control-manifest.md.
Great for
- Game studios or solo developers using the Claude Code Game Studios workflow who already have accepted ADRs
- Technical leads who want a cheat sheet for programmers derived from architecture decisions
- Teams that need the manifest regenerated after an ADR is accepted or revised
Skip it if
- Projects with no architecture decision records: the skill stops and asks you to create ADRs first
- Projects whose ADRs have no Status sections, since acceptance cannot be determined and the skill stops
- Non-game or non-engine projects, because it expects engine reference files and the studio's specific folder layout
Try asking
- “Create a control manifest from the accepted ADRs for this project”
- “Regenerate the control manifest now that I have accepted the new rendering ADR”
- “Give me programmer rules and a cheat sheet from the ADRs, with a lean review mode”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
create-control-manifest.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip create-control-manifest.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip create-control-manifest.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Run /architecture-review and accept the ADRs first, so the manifest is built from settled decisions rather than Proposed ones
- Read the preview carefully: it lists every ADR left out and the rules that could not be scoped, so you approve knowing what is missing
- Re-run the skill after each ADR change, because the manifest is replaced wholesale on each regeneration
What's inside
- SKILL.md
Redistributed unmodified from Donchitos/Claude-Code-Game-Studios under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.