steve2
Steve — wavybaby + GSD + Ralph Autonomous Execution
Combines three systems:
- wavybaby: Self-equipping toolchain (CoVe verification, skill discovery via skills.sh, MCP server setup, project config)
- GSD: Structured project management (
.planning/, roadmaps, phases, plans, discovery, verification, state tracking) - Ralph: Autonomous while-true loop (circuit breakers, dual-condition exit, session persistence, rate limiting)
IMPORTANT — Context Management: Steve offloads heavy phases to subagents via the Task tool to avoid hitting the 20MB API request limit. Phase instructions live in template files at ~/.claude/skills/steve/templates/. The main conversation stays lean — it orchestrates, collects brief summaries, and handles user interaction only.
EXECUTION FLOW
Step 1: Parallel Kickoff
Do these TWO things simultaneously in a single message with parallel tool calls:
A) Spawn Phase 0 subagent (Equip):
More from aaarnv/claude-skills
spidey
Sets up a Ralph autonomous development loop for any project. First generates a full PRD from the user's description, then derives a task plan from it. Wraps Claude Code in an intelligent while-true loop with circuit breakers, exit detection, session persistence, and progress tracking. Use when you want Claude to autonomously work through a task list until done.
86rnv
Chain-of-Verification (CoVe) prompting system. Converts lazy prompts into rigorous 4-stage verified output. Use for any code generation, debugging, or implementation task. Automatically invoked by wavybaby for medium/high complexity tasks. Reduces hallucinations and catches subtle bugs.
82steve
The ultimate autonomous dev pipeline. Combines wavybaby (CoVe verification, skill discovery, MCP tooling) + GSD (roadmaps, phases, plans, discovery, state tracking) + Ralph (autonomous loop with circuit breakers). Generates a PRD, equips itself with the best tools, bootstraps a full GSD .planning/ structure, then runs Ralph to autonomously execute each plan with CoVe-verified code until the milestone is complete.
73wavybaby
Extends Claude Code's built-in skill dispatch with CoVe (Chain-of-Verification), dynamic skill discovery via skills.sh, and a toolkit knowledge base for MCP servers and configurations Claude doesn't natively know about. Use for any non-trivial task.
64herobrine
Create, list, remove, and run scheduled autonomous Claude Code agents. Agents run on a timer via macOS launchd, execute any prompt headlessly, and deliver results via Beeper messages and macOS notifications. Use for recurring research, monitoring, overnight builds, or any task you want Claude to do on autopilot.
27