agentrigs
+ submit rig
~ / rigs / th3vib3coder / vibe-science

th3vib3coder/vibe-science

Adversarial agent loops for verifiable vibe researching. Claude Code plugin. Falsification-first, not production-first.

↗ GitHub ★ 15 apache-2.0 updated 4mo ago project Claude CodeCodex Claude plugin
share on X
ARCHETYPE
YOLO Cowboy
Permissions? Never heard of 'em. Ships at the speed of `--dangerously-skip-permissions`.
CONTEXT TAX · EVERY TURN
~5.9k tokens
Moderate · median rig: 2.3k · breakdown
GUARDRAILS
1/5
Pre-tool screening hook · details

Copy this rig

# review before running: this installs third-party code
$ npx degit th3vib3coder/vibe-science/.claude ./rig-vibe-science  # 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 th3vib3coder/vibe-science/.claude/skills/delivery-discipline .claude/skills/delivery-discipline
$ npx degit th3vib3coder/vibe-science/skills/vibe .claude/skills/vibe
$ curl -fsSL --create-dirs -o .claude/agents/reviewer2.md https://raw.githubusercontent.com/th3vib3coder/vibe-science/main/agents/reviewer2.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": [
      "Edit(**/schemas/*)",
      "Write(**/schemas/*)",
      "Edit(**/fault-taxonomy.yaml)",
      "Write(**/fault-taxonomy.yaml)",
      "Edit(**/judge-rubric.yaml)",
      "Write(**/judge-rubric.yaml)",
      "Edit(**/delivery-attestation.schema.json)",
      "Write(**/delivery-attestation.schema.json)",
      "Edit(**/plugin/scripts/pre-delivery-discipline.js)",
      "Write(**/plugin/scripts/pre-delivery-discipline.js)",
      "Edit(**/plugin/scripts/pre-tool-use.js)",
      "Write(**/plugin/scripts/pre-tool-use.js)",
      "Edit(**/.claude/settings.json)",
      "Write(**/.claude/settings.json)",
      "Edit(**/.claude/settings.local.json)",
      "Write(**/.claude/settings.local.json)",
      "Edit(**/hooks/hooks.json)",
      "Write(**/hooks/hooks.json)",
      "Edit(**/package.json)",
      "Write(**/package.json)",
      "Edit(**/package-lock.json)",
      "Write(**/package-lock.json)",
      "Edit(**/.github/workflows/ci.yml)",
      "Write(**/.github/workflows/ci.yml)",
      "Edit(**/__test_e2e.mjs)",
      "Write(**/__test_e2e.mjs)",
      "Edit(**/tests/governance-events.test.mjs)",
      "Write(**/tests/governance-events.test.mjs)",
      "Edit(**/tests/governance-hooks.test.mjs)",
      "Write(**/tests/governance-hooks.test.mjs)",
      "Edit(**/tests/delivery-discipline-skill.test.mjs)",
      "Write(**/tests/delivery-discipline-skill.test.mjs)",
      "Edit(**/tests/validate-delivery-honesty.test.mjs)",
      "Write(**/tests/validate-delivery-honesty.test.mjs)",
      "Edit(**/tests/delivery-discipline-hook.test.mjs)",
      "Write(**/tests/delivery-discipline-hook.test.mjs)",
      "Edit(**/plugin/scripts/handshake-inject.js)",
      "Write(**/plugin/scripts/handshake-inject.js)",
      "Edit(**/plugin/scripts/objective-loader.js)",
      "Write(**/plugin/scripts/objective-loader.js)",
      "Edit(**/plugin/scripts/loop-wake.js)",
      "Write(**/plugin/scripts/loop-wake.js)",
      "Edit(**/plugin/scripts/r2-bridge-writer.js)",
      "Write(**/plugin/scripts/r2-bridge-writer.js)",
      "Edit(**/vibe-research-environment/environment/orchestrator/autonomy-runtime.js)",
      "Write(**/vibe-research-environment/environment/orchestrator/autonomy-runtime.js)"
    ]
  },
  "hooks": {
    "PreToolUse": [
      {
        "matcher": "Write|Edit|MultiEdit|Bash",
        "hooks": [
          {
            "type": "command",
            "command": "node \"$CLAUDE_PROJECT_DIR/plugin/scripts/pre-tool-use.js\""
          }
        ]
      },
      {
        "matcher": "Write|Edit|MultiEdit",
        "hooks": [
          {
            "type": "command",
            "command": "node \"$CLAUDE_PROJECT_DIR/plugin/scripts/pre-delivery-discipline.js\""
          }
        ]
      }
    ]
  }
}

Skills (2)

delivery-disciplinevibe-science

Subagents (1)

reviewer2
model: opus
Adversarial scientific reviewer (Reviewer 2). Spawned as subagent to DESTROY weak claims. Assumes every claim is wrong. Demands counter-evidence, confounder harness, and falsification tests. Must sear

Hooks (8)

eventmatcherruns
SessionStart*node "$CLAUDE_PROJECT_DIR/plugin/scripts/session-start.js"
UserPromptSubmit*node "$CLAUDE_PROJECT_DIR/plugin/scripts/prompt-submit.js"
PostToolUse*node "$CLAUDE_PROJECT_DIR/plugin/scripts/post-tool-use.js"
Stop*node "$CLAUDE_PROJECT_DIR/plugin/scripts/stop.js"
PreCompact*node "$CLAUDE_PROJECT_DIR/plugin/scripts/pre-compact.js"
PreToolUseWrite|Edit|MultiEdit|Bashnode "$CLAUDE_PROJECT_DIR/plugin/scripts/pre-tool-use.js"
PreToolUseWrite|Edit|MultiEditnode "$CLAUDE_PROJECT_DIR/plugin/scripts/pre-delivery-discipline.js"
SubagentStop*node "$CLAUDE_PROJECT_DIR/plugin/scripts/subagent-stop.js"

Slash commands (9)

/atomic-check/ingest-source/init/lint-wiki/loop/query/reviewer2/search/start

Permissions

deny (46)
Edit(**/schemas/*)
Write(**/schemas/*)
Edit(**/fault-taxonomy.yaml)
Write(**/fault-taxonomy.yaml)
Edit(**/judge-rubric.yaml)
Write(**/judge-rubric.yaml)
Edit(**/delivery-attestation.schema.json)
Write(**/delivery-attestation.schema.json)
Edit(**/plugin/scripts/pre-delivery-discipline.js)
Write(**/plugin/scripts/pre-delivery-discipline.js)
Edit(**/plugin/scripts/pre-tool-use.js)
Write(**/plugin/scripts/pre-tool-use.js)
Edit(**/.claude/settings.json)
Write(**/.claude/settings.json)
Edit(**/.claude/settings.local.json)
Write(**/.claude/settings.local.json)
Edit(**/hooks/hooks.json)
Write(**/hooks/hooks.json)
Edit(**/package.json)
Write(**/package.json)
Edit(**/package-lock.json)
Write(**/package-lock.json)
Edit(**/.github/workflows/ci.yml)
Write(**/.github/workflows/ci.yml)
Edit(**/__test_e2e.mjs)
Write(**/__test_e2e.mjs)
Edit(**/tests/governance-events.test.mjs)
Write(**/tests/governance-events.test.mjs)
Edit(**/tests/governance-hooks.test.mjs)
Write(**/tests/governance-hooks.test.mjs)
Edit(**/tests/delivery-discipline-skill.test.mjs)
Write(**/tests/delivery-discipline-skill.test.mjs)
Edit(**/tests/validate-delivery-honesty.test.mjs)
Write(**/tests/validate-delivery-honesty.test.mjs)
Edit(**/tests/delivery-discipline-hook.test.mjs)
Write(**/tests/delivery-discipline-hook.test.mjs)
Edit(**/plugin/scripts/handshake-inject.js)
Write(**/plugin/scripts/handshake-inject.js)
Edit(**/plugin/scripts/objective-loader.js)
Write(**/plugin/scripts/objective-loader.js)
Edit(**/plugin/scripts/loop-wake.js)
Write(**/plugin/scripts/loop-wake.js)
Edit(**/plugin/scripts/r2-bridge-writer.js)
Write(**/plugin/scripts/r2-bridge-writer.js)
Edit(**/vibe-research-environment/environment/orchestrator/autonomy-runtime.js)
Write(**/vibe-research-environment/environment/orchestrator/autonomy-runtime.js)
ask (0)
—
allow (17)
Read
Edit
Write
Bash
Glob
Grep
Agent
WebSearch
WebFetch
Skill
mcp__*
Bash(*)
Read(*)
Write(*)
Edit(*)
Glob(*)
Grep(*)

Similar rigs

shanraisshan/claude-code-best-practice
from vibe coding to agentic engineering - practice makes claude perfect
★ 67.0k project Claude CodeCodex 3 mcp 6 skills 20 agents 30 hooks 7d ago
YOLO Cowboy 12.9k tok ·
davila7/claude-code-templates
CLI tool for configuring and monitoring Claude Code
★ 32.3k project Claude CodeCodex 27 mcp 3 skills 15 agents 21 hooks 7d ago
YOLO Cowboy 80.9k tok ·
IvanMurzak/Unity-MCP
AI Skills, MCP Tools, and CLI for Unity Engine. Full AI develop and test loop. Use cli for quick setup. Efficient token usage, advanced tools. Any C# method may be turned into a tool by a single line. Works with Claude Code, Gemini, Copilot
★ 4,394 project Claude CodeCopilot 80 skills 5d ago
YOLO Cowboy 5.9k tok ·
MaxMiksa/Auto-Company
An auto-company works for 24/7 on your own PC - Windows/Linux/macOS.
★ 3,112 project Claude Code 36 skills 28 agents 6d ago
YOLO Cowboy 7.8k tok ·
copied ✓