~ / rigs / o8-run / o8

o8-run/o8

The review and merge layer for coding agents. Your agents write the code; o8 reviews it, sends it back until it's right, and merges it. Works with Claude Code, Codex, T3 Code and more.

↗ GitHub ★ 50 MIT updated 4d ago project Claude CodeCodexGemini CLI
share on X
ARCHETYPE
Skill Collector
Ten-plus skills loaded on demand. A procedural-knowledge library.
CONTEXT TAX · EVERY TURN
~4.9k tokens
Moderate · 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
$ claude mcp add o8 -- npx tsx src/lib/mcp/operator-mcp-server.ts
$ npx degit o8-run/o8/.claude ./rig-o8  # 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.

$ claude mcp add o8 -- npx tsx src/lib/mcp/operator-mcp-server.ts
$ npx degit o8-run/o8/.agents/skills/o8-checks .claude/skills/o8-checks
$ npx degit o8-run/o8/.agents/skills/o8-code .claude/skills/o8-code
$ npx degit o8-run/o8/.agents/skills/o8-contribute .claude/skills/o8-contribute
$ npx degit o8-run/o8/.agents/skills/o8-delegate .claude/skills/o8-delegate
$ npx degit o8-run/o8/.agents/skills/o8-release .claude/skills/o8-release
$ npx degit o8-run/o8/.agents/skills/o8-ui .claude/skills/o8-ui
$ npx degit o8-run/o8/.agents/skills/ship .claude/skills/ship
$ npx degit o8-run/o8/.claude/skills/o8-dev .claude/skills/o8-dev
$ npx degit o8-run/o8/.claude/skills/reports .claude/skills/reports
$ npx degit o8-run/o8/plugins/o8/skills/check-status .claude/skills/check-status
$ npx degit o8-run/o8/plugins/o8/skills/connected-o8 .claude/skills/connected-o8
$ npx degit o8-run/o8/plugins/o8/skills/hand-off-work .claude/skills/hand-off-work
$ curl -fsSL --create-dirs -o .claude/agents/dogfood-audit.md https://raw.githubusercontent.com/o8-run/o8/main/.claude/agents/dogfood-audit.md
$ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/o8-run/o8/main/.claude/agents/reviewer.md
$ curl -fsSL --create-dirs -o .claude/agents/runtime-debugger.md https://raw.githubusercontent.com/o8-run/o8/main/.claude/agents/runtime-debugger.md
$ curl -fsSL --create-dirs -o .claude/agents/smoke-tester.md https://raw.githubusercontent.com/o8-run/o8/main/.claude/agents/smoke-tester.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 *)"
    ]
  }
}

MCP servers (1)

serversourceest. tokens
tsx · "o8" npm 2.5k

Skills (12)

Subagents (4)

dogfood-audit
model: opus
Full product audit — spawns 5-6 parallel agents to dogfood every surface of o8, checking orchestrator, agents, terminals, diffs, UI via Playwright, and reporting issues
reviewer
model: sonnet
Use this agent PROACTIVELY after code changes to check for CLAUDE.md rule violations, render loops, security issues, and React anti-patterns. Read-only — never writes code.
runtime-debugger
model: sonnet
Use this agent PROACTIVELY when debugging runtime adapter issues — session discovery failures, WebSocket RPC errors, gateway communication problems, stale/ghost agents, or agent lifecycle bugs.
smoke-tester
model: sonnet
Use this agent PROACTIVELY after making multiple changes, before committing, or when the user asks "does everything work." Checks dev server health, API endpoints, type safety, and project hygiene.

Hooks (1)

eventmatcherruns
PostToolUseStop|TaskCompletenode "${CLAUDE_PROJECT_DIR:-$(git rev-parse --show-toplevel)}/dist/hooks/completion-gate.js"

Similar rigs

copied ✓