~ / rigs / tigran-martirosian / ai-coding-setup

tigran-martirosian/ai-coding-setup

Claude Code setup: hooks check the agent's replies against its action log and send back claims it doesn't show. A planner splits big jobs, reviewers check them, and Nimbalyst shows progress and usage. Node.js, TypeScript, React; 116 test fi

↗ GitHub ★ 0 NOASSERTION updated today personal setup Claude Code
share on X
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
CONTEXT TAX · EVERY TURN
~7.1k tokens
Heavy · median rig: 2.3k · breakdown
GUARDRAILS
1/5
No YOLO mode · details
Quality 20/100 Guardrails v2

Copy this rig

# review before running: this installs third-party code
$ npx degit tigran-martirosian/ai-coding-setup/projects/claude-settings/.claude ./rig-ai-coding-setup  # 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 tigran-martirosian/ai-coding-setup/projects/quick-tasks/.claude/skills/finder .claude/skills/finder
$ npx degit tigran-martirosian/ai-coding-setup/projects/quick-tasks/.claude/skills/read-web .claude/skills/read-web
$ npx degit tigran-martirosian/ai-coding-setup/skills/backup .claude/skills/backup
$ npx degit tigran-martirosian/ai-coding-setup/skills/board-cleanup .claude/skills/board-cleanup
$ npx degit tigran-martirosian/ai-coding-setup/skills/btw .claude/skills/btw
$ npx degit tigran-martirosian/ai-coding-setup/skills/design-system .claude/skills/design-system
$ npx degit tigran-martirosian/ai-coding-setup/skills/disk-tidy .claude/skills/disk-tidy
$ npx degit tigran-martirosian/ai-coding-setup/skills/file-upkeep .claude/skills/file-upkeep
$ npx degit tigran-martirosian/ai-coding-setup/skills/grill-me .claude/skills/grill-me
$ npx degit tigran-martirosian/ai-coding-setup/skills/handoff .claude/skills/handoff
$ npx degit tigran-martirosian/ai-coding-setup/skills/lead .claude/skills/lead
$ npx degit tigran-martirosian/ai-coding-setup/skills/new-project .claude/skills/new-project
$ npx degit tigran-martirosian/ai-coding-setup/skills/next-move .claude/skills/next-move
$ npx degit tigran-martirosian/ai-coding-setup/skills/pc-tidy .claude/skills/pc-tidy
$ npx degit tigran-martirosian/ai-coding-setup/skills/phone-bot .claude/skills/phone-bot
$ npx degit tigran-martirosian/ai-coding-setup/skills/setup-audit .claude/skills/setup-audit
$ npx degit tigran-martirosian/ai-coding-setup/skills/site-login .claude/skills/site-login
$ npx degit tigran-martirosian/ai-coding-setup/skills/systems-map .claude/skills/systems-map
$ npx degit tigran-martirosian/ai-coding-setup/skills/untangle-codebase .claude/skills/untangle-codebase
$ npx degit tigran-martirosian/ai-coding-setup/skills/update-project .claude/skills/update-project
$ npx degit tigran-martirosian/ai-coding-setup/skills/update-setup .claude/skills/update-setup
$ npx degit tigran-martirosian/ai-coding-setup/skills/watch .claude/skills/watch
$ npx degit tigran-martirosian/ai-coding-setup/skills/web-research .claude/skills/web-research
$ npx degit tigran-martirosian/ai-coding-setup/skills/where-are-we .claude/skills/where-are-we
$ npx degit tigran-martirosian/ai-coding-setup/skills/writing .claude/skills/writing
$ curl -fsSL --create-dirs -o .claude/agents/blind-reviewer.md https://raw.githubusercontent.com/tigran-martirosian/ai-coding-setup/main/agents/blind-reviewer.md
$ curl -fsSL --create-dirs -o .claude/agents/plan-reviewer.md https://raw.githubusercontent.com/tigran-martirosian/ai-coding-setup/main/agents/plan-reviewer.md
$ curl -fsSL --create-dirs -o .claude/agents/planner.md https://raw.githubusercontent.com/tigran-martirosian/ai-coding-setup/main/agents/planner.md
$ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/tigran-martirosian/ai-coding-setup/main/agents/reviewer.md
$ curl -fsSL --create-dirs -o .claude/agents/stuck-advisor.md https://raw.githubusercontent.com/tigran-martirosian/ai-coding-setup/main/agents/stuck-advisor.md
$ curl -fsSL --create-dirs -o .claude/agents/worker-code.md https://raw.githubusercontent.com/tigran-martirosian/ai-coding-setup/main/agents/worker-code.md
$ curl -fsSL --create-dirs -o .claude/agents/worker-full.md https://raw.githubusercontent.com/tigran-martirosian/ai-coding-setup/main/agents/worker-full.md
$ curl -fsSL --create-dirs -o .claude/agents/worker-refactor.md https://raw.githubusercontent.com/tigran-martirosian/ai-coding-setup/main/agents/worker-refactor.md
$ curl -fsSL --create-dirs -o .claude/agents/worker.md https://raw.githubusercontent.com/tigran-martirosian/ai-coding-setup/main/agents/worker.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 (25)

Subagents (9)

blind-reviewer
model: fable
A blind look at something a person will see, read or use, judged as a newcomer meeting it for the first time, with no code, plan or reasoning. Use for a mock-up, first screen or draft before the build
plan-reviewer
model: fable
One independent read-only pass over a written plan and its job-list rows before the user sees it, against the user's words and the decisions, and for whether each row can run as written. Runs the plan
planner
model: fable
Writes a plan in two passes on Fable at high effort - first a one-page approach for the user's OK, then after that OK the job-list rows (each step's helper, model, budget, what it decides, and whether
reviewer
model: fable
One independent read-only pass over a finished bigger change (a job of several steps, 10 or more code files, or hooks and settings), against the request and the plan and for quality. One for a job; 2
stuck-advisor
model: fable
A second opinion from Fable on a problem the session is stuck on, while the session keeps doing the work. Call it when the same error comes back after 2 honest fixes; when a choice the plan did not se
worker-code
model: opus
Step worker for code changes, on Opus at high effort. Use it for a step or one-off job that writes or changes code. A long refactor that is not mechanical, with dependents across files, goes to worker
worker-full
model: opus
Same step worker on Opus at high effort, with every tool and the CLAUDE.md files loaded. Costs about 34k more tokens to start, so only for a job that needs a skill, an MCP tool (browser, Nimbalyst) or
worker-refactor
model: fable
Step worker on Fable at high effort, only for a long refactor step that is not mechanical. Use it when the step changes something code in other files depends on (an interface, signature, type or data
worker
model: opus
Step worker. Does one numbered step of a job the planner runs, knowing the job's goal, or one clear one-off job - repetitive edits, renames, formatting, drafting docs or tests, web gathering. Starts w

Hooks (1)

eventmatcherruns
Stop*node .claude/hooks/finder-link-gate.mjs

Similar rigs

copied ✓