MysticalNobody/alatyr_flutter
A hardened Flutter starter and GitHub template for AI-agent development, with machine-enforced architecture, automated quality gates, and a Claude Code + Codex implementation/review workflow
ARCHETYPE
Pragmatist
A balanced, no-drama setup: some rules, some tools, nothing extreme.
Copy this rig
# review before running: this installs third-party code
$ npx degit MysticalNobody/alatyr_flutter/.claude ./rig-alatyr_flutter # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit MysticalNobody/alatyr_flutter/.agents/skills/cross-review .claude/skills/cross-review $ npx degit MysticalNobody/alatyr_flutter/.claude/skills/adversarial-tests .claude/skills/adversarial-tests
$ curl -fsSL --create-dirs -o .claude/agents/codex-reviewer.md https://raw.githubusercontent.com/MysticalNobody/alatyr_flutter/main/.claude/agents/codex-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/test-breaker.md https://raw.githubusercontent.com/MysticalNobody/alatyr_flutter/main/.claude/agents/test-breaker.md
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"permissions": {
"deny": [
"Read(/.dart-defines/*.env)"
]
},
"hooks": {
"PreToolUse": [
{
"matcher": "Edit|Write",
"hooks": [
{
"type": "command",
"command": "\"$CLAUDE_PROJECT_DIR\"/tool/hooks/guard_generated.sh"
}
]
}
]
}
} Skills (2)
Subagents (2)
| codex-reviewer model: sonnet | Thin wrapper that runs the Codex cross-review script and returns its output verbatim - for workflows and skills that need the review as a node; it never edits files and never interprets findings. |
| test-breaker model: sonnet | Fresh-context adversarial scenario generator - given a feature's *_api contract, its spec/plan text and the diff, lists the ways the implementation could break (boundary values, races and double taps, |
Hooks (2)
| event | matcher | runs |
|---|---|---|
| PreToolUse | Edit|Write | "$CLAUDE_PROJECT_DIR"/tool/hooks/guard_generated.sh |
| PostToolUse | Edit|Write | "$CLAUDE_PROJECT_DIR"/tool/hooks/format_dart.sh |
Permissions
deny (1)
Read(/.dart-defines/*.env)
ask (0)
—
allow (21)
Bash(fvm flutter analyze:*)
Bash(fvm flutter test:*)
Bash(fvm flutter pub get:*)
Bash(fvm dart analyze:*)
Bash(fvm dart test:*)
Bash(fvm dart format:*)
Bash(fvm dart run:*)
Bash(fvm dart pub get:*)
Bash(flutter analyze:*)
Bash(flutter test:*)
Bash(dart analyze:*)
Bash(dart test:*)
Bash(dart format:*)
Bash(dart run:*)
Bash(tool/checks.sh:*)
Bash(tool/codegen.sh:*)
Bash(tool/e2e.sh:*)
Bash(tool/template_smoke.sh:*)
Bash(tool/web_smoke.sh:*)
Bash(bash tool/checks.sh:*)
Bash(bash tool/codegen.sh:*)
Similar rigs
cjcsecurity/claude-code-dual-build
Symmetric multi-agent build with mandatory cross-review for Claude Code + Codex. Splits coding tasks ~50/50, runs in parallel in isolated worktrees, has the opposite model review each diff.
Pragmatist 429 tok ·
witekbobrowski/dotfiles
🗂 Automated custom macOS config
Orchestrator 988 tok ·
hirokisakabe/issuekit
Agent Skills for issue-driven development.
Skill Collector 5.6k tok ·
Sateezg/codex-bridge
Image generation (gpt-image-2) and GPT-5 subagents for Claude Code — through the Codex CLI login you already have. No OpenAI API key.
Orchestrator 946 tok ·