Claude Code Vertical

Scale your Claude Code usage horizontally and vertically.

Horizontal: Run multiple planning sessions in parallel

Vertical: Each plan spawns multiple weavers executing specs concurrently

Installation

curl -fsSL https://raw.githubusercontent.com/harivansh-afk/claude-code-vertical/main/install.sh | bash

Requires: Claude Code CLI, tmux (brew install tmux)

Quick Start

# Start a planning session
claude
> /plan

# Design specs interactively with the planner...
# When ready:
> /build plan-20260119-1430

# Check status
> /status plan-20260119-1430

Documentation

Document	Description
WORKFLOW.md	Complete workflow guide with examples
CLAUDE.md	Project instructions and quick reference
docs/spec-schema-v2.md	Full spec YAML schema

Architecture

You (Terminal)
    │
    │ /plan
    ▼
Planner (interactive)     ← You talk here
    │
    │ (writes specs)
    ▼
Orchestrator (tmux background)
    │
    ├─→ Weaver 01 (tmux) → Verifier (subagent) → PR
    ├─→ Weaver 02 (tmux) → Verifier (subagent) → PR
    └─→ Weaver 03 (tmux) → Verifier (subagent) → PR

Directory Structure

claude-code-vertical/
├── CLAUDE.md              # Project instructions
├── WORKFLOW.md            # Complete workflow guide
├── skills/
│   ├── planner/           # Interactive planning
│   ├── orchestrator/      # Tmux + weaver management
│   ├── weaver-base/       # Base skill for all weavers
│   ├── verifier/          # Verification subagent
│   └── oracle/            # Deep planning with GPT-5.2 Codex
├── skill-index/
│   ├── index.yaml         # Skill registry
│   └── skills/            # Available skills
├── lib/
│   └── tmux.sh            # Tmux helper functions
└── .claude/
    └── vertical/
        └── plans/         # Your plans live here

All Agents Use Opus

Every agent uses claude-opus-4-5-20250514 for maximum capability.

Key Rules

1. Tests never ship - Weavers may write tests for verification, but they're never committed
2. PRs always created - Weaver success = PR created
3. Verification mandatory - Weavers spawn verifier subagents
4. Context isolated - Each agent sees only what it needs

Skill Index

The orchestrator matches skill_hints from specs to skills in skill-index/index.yaml.

Resume Any Session

# Find session ID
cat .claude/vertical/plans/<plan-id>/run/weavers/w-01.json | jq -r .session_id

# Resume
claude --resume <session-id>

Tmux Helpers

source lib/tmux.sh

vertical_status              # Show all plans
vertical_list_sessions       # List tmux sessions
vertical_attach <session>    # Attach to session
vertical_kill_plan <plan-id> # Kill all sessions for a plan

Oracle for Complex Planning

For complex tasks, the planner invokes Oracle (GPT-5.2 Codex) for deep planning:

npx -y @steipete/oracle --engine browser --model gpt-5.2-codex ...

Oracle runs 10-60 minutes and outputs plan.md, which the planner transforms into specs.