Deploy 16 to 128 AI agents in parallel from a single terminal. You type objectives. The system deploys a coordinated fleet.
Nine sequential phases. Each fans out into parallel agents, then collapses into a shared artifact before the next phase begins.
Strategic agents. Read-only. Never write code. See the big picture, make architectural decisions.
Implementation agents. Full permissions. The workhorse. Deep execution on focused tasks.
Fast scouts. Read-only. Quick in, quick out. Cheap and expendable. Use liberally.
Deploy N features in parallel with full test coverage. The flagship OPBLITZ pattern.
Full security sweep. OWASP checks, dependency audit, attack surface mapping.
Migrate between frameworks or versions. Scan, transform, verify per module.
Write comprehensive tests for the entire codebase. One tester per module.
Refactor N modules simultaneously. Identify smells, transform, verify.
FAST TRACK. Diagnose + fix production incidents. 8 scouts simultaneous.
Document everything in parallel. One writer per module.
Research N topics in parallel, synthesize findings into one report.
The first field deployment. One operator, 16 agents, 4 features, 13 minutes.
Open source. Zero dependencies beyond Claude Code. Copy the agents, pick an operation, deploy.