Superpowers

What Superpowers does

Superpowers packages a software development workflow as a set of composable skills plus initial instructions that make the agent use them. When the agent sees that you are building something, it first asks what you are trying to do, then presents a spec in chunks short enough to read. After you sign off on the design, it writes an implementation plan that stresses red/green TDD, YAGNI and DRY.

Once you say go, it starts a subagent-driven-development process in which agents work through each engineering task while their output is inspected and reviewed. The README says an agent can work autonomously for a couple of hours without deviating from the plan. Because the skills trigger automatically, no special commands are needed. Installation differs by harness, so Superpowers is installed separately for each tool you use, and the README has a dedicated Hermes Agent section.

Key features

• Asks clarifying questions before writing code and shows the spec in short chunks
• Writes implementation plans built around red/green TDD, YAGNI and DRY
• Runs subagent-driven development with review of each engineering task
• Skills trigger automatically without special commands
• Separate install sections for Claude Code, Codex, Cursor, Gemini CLI, OpenCode, Hermes Agent and others

When to use it

• Turning a vague feature request into an approved spec and plan before any code is written
• Letting an agent work through a multi-task plan for a couple of hours with review between tasks
• Applying one development methodology across several coding agents

Who it is for: Developers who use coding agents and want a structured plan, test and review workflow instead of ad hoc prompting.

How it fits with Hermes Agent

Superpowers is a general framework for many coding agents; Hermes Agent is one of the harnesses with its own section in the installation guide.

FAQ