Agentic skills framework and software development methodology for Claude Code. 30K stars, 2K forks, 13 contributors, 64 open issues.
Key Features (from video)
- Code Review — automated code review as a skill
- Security Guidance — installable via
claude /install anthropics/security-guidance. Scans for: command injection, eval()/new Function(), XSS vectors, unsafe deserialization, os.system() injection, path traversal, SQL injection, hardcoded secrets - Full dev cycle: planning, subagents, TDD, code review
Status
Previously evaluated 2026-04-08 and skipped — 22K tokens at startup (11% of context), forces full brainstorm-TDD cycle on every task regardless of size. A custom /implement-spec skill covers the core value without the overhead.
New signal: The security-guidance plugin is a standalone install separate from the full superpowers framework. Worth installing independently for projects where code handles real operations data.
Action
- Install
anthropics/security-guidancestandalone — test on current projects - Re-evaluate superpowers if context window increases or token cost drops
Source
- Platform: Instagram Reel
- Author: @shishkin
- Repo: obra/superpowers (GitHub)
- Date: 2026-04-12
part of tooling