saadnvd1/agent-os
Mobile-first web UI for managing AI coding sessions (Claude Code, Codex, Aider, Gemini CLI, Amp, Pi). Self-hosted with multi-pane terminals, git integration, and session orchestration.
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code
$ npx degit saadnvd1/agent-os/.claude ./rig-agent-os # 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 saadnvd1/agent-os/.claude/skills/debugging .claude/skills/debugging $ npx degit saadnvd1/agent-os/.claude/skills/do-code-review .claude/skills/do-code-review
$ curl -fsSL --create-dirs -o .claude/agents/review-durability.md https://raw.githubusercontent.com/saadnvd1/agent-os/main/.claude/agents/review-durability.md $ curl -fsSL --create-dirs -o .claude/agents/review-finding-verifier.md https://raw.githubusercontent.com/saadnvd1/agent-os/main/.claude/agents/review-finding-verifier.md $ curl -fsSL --create-dirs -o .claude/agents/review-frontend.md https://raw.githubusercontent.com/saadnvd1/agent-os/main/.claude/agents/review-frontend.md $ curl -fsSL --create-dirs -o .claude/agents/review-orchestrator-safety.md https://raw.githubusercontent.com/saadnvd1/agent-os/main/.claude/agents/review-orchestrator-safety.md $ curl -fsSL --create-dirs -o .claude/agents/review-security.md https://raw.githubusercontent.com/saadnvd1/agent-os/main/.claude/agents/review-security.md $ curl -fsSL --create-dirs -o .claude/agents/review-spec-compliance.md https://raw.githubusercontent.com/saadnvd1/agent-os/main/.claude/agents/review-spec-compliance.md $ curl -fsSL --create-dirs -o .claude/agents/review-test-quality.md https://raw.githubusercontent.com/saadnvd1/agent-os/main/.claude/agents/review-test-quality.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 7,204 rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(rm -rf *)",
"Read(**/*.pem)",
"Bash(rm -rf /)",
"Bash(git push --force:*)",
"Bash(sudo *)",
"Read(.env)",
"Bash(rm -rf /*)",
"Read(./.env.*)",
"Read(~/.aws/**)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(sudo:*)",
"Bash(git reset --hard*)",
"Bash(git reset --hard:*)",
"Read(.env.*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(rm:*)",
"Bash(git rebase *)",
"Bash(wget *)",
"Bash(npm publish:*)",
"Bash(git commit *)",
"Bash(gh pr merge *)"
]
}
} Skills (2)
Subagents (7)
| review-durability model: inherit | Review AgentOS changes for lost or duplicated work across restarts - the chat worker protocol, restart and reattach, idempotent sends, migrations appended with the next id, state claimed before side e |
| review-finding-verifier model: inherit | Independently confirm or reject findings from the other review agents by reading the code. Dispatched by the do-code-review skill after the reviewers report, with up to 5 findings per call. |
| review-frontend model: inherit | Review AgentOS UI changes - phone-first layouts with 44px touch targets, light and dark themes both working, React Query patterns in data/, no AI-slop visuals, and accessibility basics. Dispatched by |
| review-orchestrator-safety model: inherit | Review changes to the orchestrator, tasks and stacks for merges and approvals that could happen when they shouldn't - gates failing open, merges not pinned to the reviewed commit, workspace scoping, b |
| review-security model: inherit | Review AgentOS changes for security holes - the device-token gate, loopback and tailnet trust, forwarding headers, Host/Origin and DNS rebinding, WebSocket upgrades, pairing and passkeys, secrets in l |
| review-spec-compliance model: inherit | Check a branch against what it was asked to do (the AgentOS task prompt, the PR body, the LumifyHub card, the plan it points at) - requirements missed and changes nobody asked for. Dispatched by the d |
| review-test-quality model: inherit | Review AgentOS test changes for tests that don't test the behaviour - missing tests for new routes, gates, parsers and tools, tests that pass for the wrong reason, over-mocking, and the NODE_ENV, CI-v |
Slash commands (2)
/commit/setup
Similar rigs
grandheman/claude-sdlc
AI-Native Software Development Lifecycle for Claude Code — 15 role-based skills covering every SDLC phase, from project kickoff to production monitoring. Works with superpowers plugin, OpenRouter council, and Playwright MCP.
Minimalist 105 tok ·
Anmolnoor/claude-dotfiles
Portable Claude Code setup — clone on a new machine, run ./install.sh, done.
Pragmatist 1.3k tok ·
ezintz/dotfiles-claude-plugin
Claude Code plugin: an env guard that makes destructive commands against non-local targets ask first, plus review, debugging and authoring skills.
Pragmatist 405 tok ·
wondping0/wondev
Write your AI agent knowledge once. Compile it to every agent config format - CLAUDE.md, AGENTS.md, .cursor/rules and more.
Pragmatist 229 tok ·