unfuck

Automate code repository cleanup by coordinating a swarm of discovery, synthesis, and implementation agents to identify issues, plan fixes, and implement changes across a multi-language codebase. This end-to-end workflow reduces manual toil and tech debt.

Yes, you can plan code cleanup changes without applying them by using the dry-run capability. This safety-first feature generates a structured cleanup plan and report without modifying the actual codebase.

Autonomous dead-code removal and codebase simplification follows a 5-phase workflow: index, discovery, synthesis, implementation, and verification. It leverages existing audit, architecture, and security skills to detect issues and implement changes safely.

Automated repo cleanup supports per-category rollbacks, LSP-based checks, and auditable commits. Safety gates are enforced throughout the implementation and verification phases to ensure codebase integrity.

Detect and fix security and architecture issues by applying an orchestrated agent swarm that audits file structure, simplifies code, and runs QA checks. It autonomously plans and implements fixes across the legacy codebase.

Yes, you can scope automated code cleanup to a specific directory. This allows you to target the discovery, synthesis, and implementation agents to a subset of the codebase rather than running end-to-end on the entire repository.