korchasa/flowai
An Assisted Engineering framework: the developer remains the architect and reviewer, while AI handles implementation under supervision.
ARCHETYPE
Skill Collector
Ten-plus skills loaded on demand. A procedural-knowledge library.
Copy this rig
# review before running: this installs third-party code
$ npx degit korchasa/flowai/.claude ./rig-flowai # 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 korchasa/flowai/.agents/skills/acceptance-tests-all .claude/skills/acceptance-tests-all $ npx degit korchasa/flowai/.agents/skills/extract-claude-code-prompt .claude/skills/extract-claude-code-prompt $ npx degit korchasa/flowai/.agents/skills/improve-primitive-from-benchmark .claude/skills/improve-primitive-from-benchmark $ npx degit korchasa/flowai/.agents/skills/root-cause-and-fix .claude/skills/root-cause-and-fix $ npx degit korchasa/flowai/.agents/skills/session-history-analyzer .claude/skills/session-history-analyzer $ npx degit korchasa/flowai/.claude/skills/diagnose-acceptance-run .claude/skills/diagnose-acceptance-run $ npx degit korchasa/flowai/.claude/skills/project-status .claude/skills/project-status $ npx degit korchasa/flowai/.claude/skills/skill-reads .claude/skills/skill-reads $ npx degit korchasa/flowai/framework/core/skills/adapt .claude/skills/adapt $ npx degit korchasa/flowai/framework/core/skills/configure-deno-commands .claude/skills/configure-deno-commands $ npx degit korchasa/flowai/framework/core/skills/epic .claude/skills/epic $ npx degit korchasa/flowai/framework/core/skills/init .claude/skills/init $ npx degit <redacted> .claude/skills/investigate $ npx degit <redacted> .claude/skills/maintenance $ npx degit korchasa/flowai/framework/core/skills/reflect-by-history .claude/skills/reflect-by-history $ npx degit korchasa/flowai/framework/core/skills/setup-ai-ide-devcontainer .claude/skills/setup-ai-ide-devcontainer $ npx degit korchasa/flowai/framework/core/skills/tasks-overview .claude/skills/tasks-overview $ npx degit korchasa/flowai/framework/core/skills/update .claude/skills/update $ npx degit korchasa/flowai/framework/core/skills/write-gods-tasks .claude/skills/write-gods-tasks $ npx degit <redacted>-skill .claude/skills/engineer-skill $ npx degit <redacted>-subagent .claude/skills/engineer-subagent $ npx degit korchasa/flowai/framework/devtools/skills/write-agent-benchmarks .claude/skills/write-agent-benchmarks $ npx degit <redacted>-research .claude/skills/deep-research $ npx degit <redacted>-benchmark-failure .claude/skills/diagnose-benchmark-failure $ npx degit <redacted>-mermaid-diagrams .claude/skills/draw-mermaid-diagrams $ npx degit <redacted>-plugin-marketplace .claude/skills/engineer-plugin-marketplace $ npx degit <redacted>-teaching-materials .claude/skills/interactive-teaching-materials $ npx degit <redacted>-dep .claude/skills/write-dep $ npx degit <redacted>-prd .claude/skills/write-prd $ npx degit korchasa/flowai/framework/swarm/skills/children-status .claude/skills/children-status $ npx degit korchasa/flowai/framework/swarm/skills/close-swarm .claude/skills/close-swarm $ npx degit korchasa/flowai/framework/swarm/skills/session-status .claude/skills/session-status $ npx degit korchasa/flowai/framework/swarm/skills/start-child .claude/skills/start-child
$ curl -fsSL --create-dirs -o .claude/agents/acceptance-test-fixer.md https://raw.githubusercontent.com/korchasa/flowai/main/.claude/agents/acceptance-test-fixer.md $ curl -fsSL --create-dirs -o .claude/agents/acceptance-test-runner.md https://raw.githubusercontent.com/korchasa/flowai/main/.claude/agents/acceptance-test-runner.md $ curl -fsSL --create-dirs -o .claude/agents/session-error-analyzer.md https://raw.githubusercontent.com/korchasa/flowai/main/.claude/agents/session-error-analyzer.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} Skills (33)
acceptance-tests-allextract-Codex-promptimprove-primitive-from-benchmarkroot-cause-and-fixsession-history-analyzerdiagnose-acceptance-runproject-statusskill-readsadaptconfigure-deno-commandsepicinitinvestigatemaintenancereflect-by-historysetup-ai-ide-devcontainertasks-overviewupdatewrite-gods-tasksengineer-skillengineer-subagentwrite-agent-benchmarksdeep-researchdiagnose-benchmark-failuredraw-mermaid-diagramsengineer-plugin-marketplaceinteractive-teaching-materialswrite-depwrite-prdchildren-statusclose-swarmsession-statusstart-child
Subagents (3)
| acceptance-test-fixer | Fixes failing acceptance test scenarios by diagnosing root causes and applying proven fix patterns. Use proactively when acceptance tests fail after skill changes, or when the user asks to fix/debug b |
| acceptance-test-runner model: inherit | Runs agent acceptance tests and interprets the results. Use proactively when you need to verify a skill or agent's performance against a scenario. |
| session-error-analyzer model: sonnet | Analyzes specific AI IDE sessions (Claude Code, OpenAI Codex, OpenCode) for LLM behavior errors and returns a structured per-session summary. Spawned by the session-history-analyzer skill with a batch |
Hooks (1)
| event | matcher | runs |
|---|---|---|
| PostToolUse | Write|Edit | deno run -A .claude/scripts/flowai-skill-structure-validate/run.ts |
Similar rigs
milosCvetkovicDev/panlode
Validate startup ideas before you build. A Claude Code plugin and workspace that scouts real customer pain, scores the evidence and kills weak ideas with cheap experiments.
Automator 2.5k tok · · Q18
yokoten-labs/yokoten
Stops the mistake you can't undo.
Skill Collector 5.9k tok · · Q20
Zidane786/plumb
Plumb: create, extract, audit and maintain a project's design system in Claude Code and Copilot CLI.
Skill Collector 6.7k tok · · Q20
tuan3w/obsidian-vault-agent
Claude Code plugin: AI agent for your Obsidian vault — process books, courses, papers, YouTube into connected notes
Skill Collector 2.7k tok · · Q0