Rompomepome/amiral
The admiral doesn't row. Orchestrator/worker model routing for Claude Code: a capable brain plans and verifies, cheap hands execute. Works inside your plan.
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 Rompomepome/amiral/.claude ./rig-amiral # 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 Rompomepome/amiral/.claude/skills/plan-ship .claude/skills/plan-ship $ curl -fsSL --create-dirs -o .claude/agents/advisor.md https://raw.githubusercontent.com/Rompomepome/amiral/main/.claude/agents/advisor.md $ curl -fsSL --create-dirs -o .claude/agents/corsaire.md https://raw.githubusercontent.com/Rompomepome/amiral/main/.claude/agents/corsaire.md $ curl -fsSL --create-dirs -o .claude/agents/grunt.md https://raw.githubusercontent.com/Rompomepome/amiral/main/.claude/agents/grunt.md $ curl -fsSL --create-dirs -o .claude/agents/implementer.md https://raw.githubusercontent.com/Rompomepome/amiral/main/.claude/agents/implementer.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/Rompomepome/amiral/main/.claude/agents/reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/advisor.md https://raw.githubusercontent.com/Rompomepome/amiral/main/agents/advisor.md $ curl -fsSL --create-dirs -o .claude/agents/corsaire.md https://raw.githubusercontent.com/Rompomepome/amiral/main/agents/corsaire.md $ curl -fsSL --create-dirs -o .claude/agents/grunt.md https://raw.githubusercontent.com/Rompomepome/amiral/main/agents/grunt.md $ curl -fsSL --create-dirs -o .claude/agents/implementer.md https://raw.githubusercontent.com/Rompomepome/amiral/main/agents/implementer.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/Rompomepome/amiral/main/agents/reviewer.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 (1)
Subagents (10)
| advisor model: opus | The expensive brain, consulted on demand. Call this agent ONLY for hard judgment calls a cheaper executor shouldn't make alone: reviewing a plan, challenging an architecture, resolving a tricky tradeo |
| corsaire model: sonnet | Licensed adversary. Pre-mortem attack on code and architecture BEFORE shipping - assumes the feature already failed in production and works backward to every cause. Use on anything risky, security-sen |
| grunt model: haiku | Mechanical, repetitive work - renames, boilerplate, mass find/replace migrations, import updates. Use for high-volume, zero-judgment tasks. |
| implementer model: sonnet | Implements a feature against a validated plan. Use whenever real code must be written across multiple files. Receives a plan, executes it, returns a summary of touched files. |
| reviewer model: sonnet | Fresh-context code review right after implementation. Use immediately after a feature has been written. Modifies nothing, returns a prioritized report. |
| advisor model: opus | The expensive brain, consulted on demand. Call this agent ONLY for hard judgment calls a cheaper executor shouldn't make alone: reviewing a plan, challenging an architecture, resolving a tricky tradeo |
| corsaire model: sonnet | Licensed adversary. Pre-mortem attack on code and architecture BEFORE shipping - assumes the feature already failed in production and works backward to every cause. Use on anything risky, security-sen |
| grunt model: haiku | Mechanical, repetitive work - renames, boilerplate, mass find/replace migrations, import updates. Use for high-volume, zero-judgment tasks. |
| implementer model: sonnet | Implements a feature against a validated plan. Use whenever real code must be written across multiple files. Receives a plan, executes it, returns a summary of touched files. |
| reviewer model: sonnet | Fresh-context code review right after implementation. Use immediately after a feature has been written. Modifies nothing, returns a prioritized report. |
Similar rigs
Saba-Var/claude-ward
A read-only tripwire that watches Claude Code's local config and reports suspicious changes.
Pragmatist 1.8k tok ·
0xT4sty/claude-code-dotfiles
—
Pragmatist 979 tok ·
OMOCHInoHOSHI/dotfiles-claude
.claude
Pragmatist 293 tok ·
jordimarsal/harness-standard
A standardized multi-agent harness for Opencode and Claude Code. Installs into any project with a single command.
Pragmatist 2.2k tok ·