beads-superpowers

English · 中文

---

A plugin for Claude Code, Codex, OpenCode, and 6 more AI coding agents that makes your agent write tests before code, debug systematically instead of guessing, and remember what it worked on yesterday. Composable skills enforce the practices; a Dolt-backed issue tracker keeps context across sessions.

How it works

When you start a task, the agent runs brainstorming to nail down requirements before touching code, then writing-plans to break the work into bd-tracked steps that survive session restarts. During implementation it follows test-driven-development (failing test first, always) and can fan out to parallel subagents via subagent-driven-development — each agent working in its own git worktree. bd stores every task, decision, and note in a local Dolt database, so the agent picks up exactly where it left off next session without relying on chat history.

Underneath all of it is a production-grade standard: the agent treats every task as if real users depend on it, so it won't quietly cut a corner, drop a requirement, or weaken a security control to move faster.

What's Inside

Testing

Skill	What it does
test-driven-development	RED-GREEN-REFACTOR loop — Iron Law: no implementation without a failing test
verification-before-completion	Evidence before claims — requires proof before marking anything done

Debugging

Skill	What it does
systematic-debugging	4-phase root-cause analysis before proposing any fix

Collaboration

Skill	What it does
requesting-code-review	Dispatches a code-reviewer subagent with structured criteria
receiving-code-review	Anti-sycophancy reception — evaluates each finding on its merits
subagent-driven-development	Fresh agent per task with spec + quality review; parallel batch mode for independent tasks
dispatching-parallel-agents	Fan-out to 2+ independent agents without shared state

Project management

Skill	What it does
brainstorming	Socratic design session before any code — produces a spec bead
stress-test	Adversarial interrogation of plans with recommended answers
writing-plans	Breaks work into bite-sized tasks, each tracked as a bd bead
executing-plans	Batch plan execution in a single session
using-git-worktrees	Isolated development branches per task
finishing-a-development-branch	Merge/PR flow + Land the Plane (close beads, push)
document-release	Post-ship doc audit — keeps README, CHANGELOG, and ARCHITECTURE in sync
project-init	Beads/Dolt DB setup, bootstrap, and recovery
tracking-with-beads	The beads conventions of record — how work gets tracked, captured, and closed out
getting-up-to-speed	Session orientation — reads the latest session-handoff doc, loads bd context, and produces a current-state summary
memory-curator	Session-close/on-demand memory consolidation — deduplicates and prunes the bd memory store
session-handoff	Human-invoked — writes a grounded handoff doc + continuation memory to resume in-progress work
research-driven-development	Parallel research agents → synthesized knowledge-base document
write-documentation	Human-quality prose — 14-rule writing system with context-first drafting

Meta

Skill	What it does
using-superpowers	Bootstrap — injected at session start, routes to the right skill

Docs

algocents.com/beads-superpowers — getting started, methodology, skills reference, example workflow, and tips.

• Example Workflow docs — Full walkthrough with diagrams
• Skills Reference — All skills explained
• Methodology — Why this workflow exists

Docs content lives in this repo (docs/ — PRs welcome); the site is built and published from a separate deployment repo.

Quickstart

The fastest path — Claude Code with native plugin install: