hydersmiling/mainseat
Opus-level results at Sonnet prices for Claude Code, with a senior engineer's discipline: minimal diffs, reuse before rewrite, self-verified changes, learned rules, secrets kept out. Sonnet builds, Haiku reads, Opus plans. Measured -57% vs
field: Developer tools
stack: Python
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 hydersmiling/mainseat/.claude ./rig-mainseat # 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 hydersmiling/mainseat/.claude/skills/compact .claude/skills/compact $ npx degit hydersmiling/mainseat/.claude/skills/lean-code .claude/skills/lean-code $ npx degit hydersmiling/mainseat/.claude/skills/learn-corrections .claude/skills/learn-corrections $ npx degit hydersmiling/mainseat/.claude/skills/map .claude/skills/map $ npx degit hydersmiling/mainseat/.claude/skills/reuse-first .claude/skills/reuse-first $ npx degit hydersmiling/mainseat/.claude/skills/safe-refactor .claude/skills/safe-refactor $ npx degit hydersmiling/mainseat/.claude/skills/self-check .claude/skills/self-check $ npx degit hydersmiling/mainseat/.claude/skills/setup .claude/skills/setup $ npx degit hydersmiling/mainseat/.claude/skills/terse .claude/skills/terse
$ curl -fsSL --create-dirs -o .claude/agents/big-decision.md https://raw.githubusercontent.com/hydersmiling/mainseat/main/agents/big-decision.md $ curl -fsSL --create-dirs -o .claude/agents/builder.md https://raw.githubusercontent.com/hydersmiling/mainseat/main/agents/builder.md $ curl -fsSL --create-dirs -o .claude/agents/debugger.md https://raw.githubusercontent.com/hydersmiling/mainseat/main/agents/debugger.md $ curl -fsSL --create-dirs -o .claude/agents/explorer.md https://raw.githubusercontent.com/hydersmiling/mainseat/main/agents/explorer.md $ curl -fsSL --create-dirs -o .claude/agents/planner.md https://raw.githubusercontent.com/hydersmiling/mainseat/main/agents/planner.md $ curl -fsSL --create-dirs -o .claude/agents/researcher.md https://raw.githubusercontent.<redacted>.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/hydersmiling/mainseat/main/agents/reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/runner.md https://raw.githubusercontent.com/hydersmiling/mainseat/main/agents/runner.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(.env)",
"Read(.env.*)",
"Read(*.pem)",
"Read(*.key)",
"Read(*.p12)",
"Read(*.pfx)",
"Read(id_rsa*)",
"Read(id_ed25519*)",
"Read(credentials*)",
"Read(secrets.*)",
"Read(.netrc)",
"Read(.npmrc)",
"Read(.pypirc)",
"Read(service-account*.json)",
"Read(node_modules/**)",
"Read(dist/**)",
"Read(build/**)",
"Read(.next/**)",
"Read(target/**)",
"Read(coverage/**)",
"Read(package-lock.json)",
"Read(pnpm-lock.yaml)",
"Read(yarn.lock)",
"Read(poetry.lock)"
]
}
} Skills (9)
Subagents (8)
| big-decision model: fable | > |
| builder model: sonnet | > |
| debugger model: sonnet | > |
| explorer model: haiku | Read-only explorer that does all code reading and searching for the main session, including single files and named symbols. Returns path:line citations, plus a short answer when asked a question; its |
| planner model: opus | > |
| researcher model: sonnet | > |
| reviewer model: haiku | > |
| runner model: haiku | > |
Permissions
deny (24)
Read(.env)
Read(.env.*)
Read(*.pem)
Read(*.key)
Read(*.p12)
Read(*.pfx)
Read(id_rsa*)
Read(id_ed25519*)
Read(credentials*)
Read(secrets.*)
Read(.netrc)
Read(.npmrc)
Read(.pypirc)
Read(service-account*.json)
Read(node_modules/**)
Read(dist/**)
Read(build/**)
Read(.next/**)
Read(target/**)
Read(coverage/**)
Read(package-lock.json)
Read(pnpm-lock.yaml)
Read(yarn.lock)
Read(poetry.lock)
ask (0)
—
allow (0)
—
Similar rigs
iuliandita/agents
Global prompts and subagent roles for coding agents, deployed to every harness you use.
Orchestrator 1.5k tok · · Q3
jaqubowsky/fleet
Runs Claude Code and pi agents in parallel on your Mac, one Docker sandbox per issue. A host agent hands out the work, watches every sandbox and brings back tested, reviewed pull requests.
Minimalist 566 tok · · Q0
Fili999/ai-hub
One hub for your AI coding agents, skills and workflows: write them once, install them into Claude Code, Codex, Antigravity CLI and local models. Free, open source, runs on your computer.
Minimalist 903 tok · · Q8
tylerjkells/claude-dotfiles
—
Pragmatist 579 tok · · Q20