Your AI security audit just got a team.

One model can't be a security expert, a performance specialist, and a documentation reviewer at the same time.

So we stopped asking it to be.

Attune AI v4.0 gives every workflow its own team of Claude subagents. A security audit now runs 4 specialists in parallel:

  attune workflow run security-audit

Behind the scenes:

  Dependency checker scans your requirements
  Injection detector reviews your code paths
  Secrets scanner catches leaked credentials
  Report writer synthesizes everything

Same command you already know. Same output format. The only difference: depth.

15 workflows now have Agent SDK adapters. Each one assembles 2-6 specialist subagents depending on the task.

The best part: zero configuration.

If you have claude-agent-sdk installed, you get the multi-agent version. If not, you get the single-model version. Attune picks the right one automatically.

  pip install 'attune-ai[developer]'
  pip install claude-agent-sdk

One install and your next run uses the full agent team.

Other things that shipped in v4.0:

  Smart workflow routing with transparent SDK/API fallback
  Deduplicated workflow listing with [SDK]/[API] tags
  Bug fixes across 18 modules
  11,800+ tests at 80% coverage

Full changelog: https://github.com/Smart-AI-Memory/attune-ai
PyPI: https://pypi.org/project/attune-ai/

What workflows do you wish had specialist AI teams?

#AIDevelopment #ClaudeAI #AgentSDK #DeveloperTools #Python #OpenSource
