AI Workflows & Agent Orchestration for Claude Code
18 multi-agent workflows for code review, security, testing, and release. 30 MCP tools. 10 auto-invoking skills. Just type /attune and go.
Security audit complete. Score: 95/100. Cost: $0.03.
Start Building in Seconds.
Install from PyPI and run your first workflow. No API key needed in Claude Code — workflows run as skills using your subscription.
- ✓Native Claude Code Plugin
- ✓30 MCP Tools Built In
- ✓Open Source — Apache 2.0
from attune.workflows import list_workflows
# See all 18 available workflows
workflows = list_workflows()
for wf in workflows:
print(f"{wf['name']:20s} stages={len(wf.get('stages', []))}")
# Run a security audit on your codebase
from attune.workflows.security_audit import SecurityAuditWorkflow
audit = SecurityAuditWorkflow()
result = await audit.execute(path="./src")
print(f"Score: {result.final_output.get('health_score')}/100")
print(f"Cost: ${result.cost_report.total_cost:.2f}")What Attune AI Does
18 Multi-Agent Workflows
Code review, security audit, test generation, release prep — each runs a specialist team of 2-6 Claude subagents with intelligent model routing.
Socratic Discovery
Workflows ask clarifying questions before executing — what scope? which files? what depth? — so you get exactly the analysis you need.
Smart Wizards
Pre-configured workflow templates for Claude Code. Deploy multi-step guided workflows — from security audits to release prep — without writing boilerplate.
Ready to automate your dev workflows?
Install from PyPI. Type /attune in Claude Code. Go.
Path traversal protection, SSRF validation, rate limiting, module import restrictions, and PreToolUse security hooks.
Opus for security, Sonnet for analysis, Haiku for scanning. Right model per task with configurable budget caps ($0.50 – $5.00).