team-dev

You can run parallel implementation tasks by spawning multiple agent implementers that work concurrently on separate plan tasks, while enforcing two-stage spec compliance and code quality review gates that block progress until both reviews pass. A lead orchestrator manages staggered reviewer dispatch as each implementer finishes.

Executing independent development tasks in parallel requires tmux-based split-pane execution, where a lead orchestrator spawns implementer teammates that work concurrently on separate file surface areas. Reviews are dispatched as soon as each implementer finishes to keep throughput high.

Parallel agent orchestration supports Jira-native workflows by executing implementation plans that contain multiple independent development tasks safely separated by file surface area. The lead orchestrator hands off deterministic plan tasks to implementer teammates and reviewer subagents.

Reuse agent teammates when subsequent parallel coding tasks are related to preserve context, and respawn clean teammates when tasks are in different areas. This staggered orchestration approach maintains high throughput while keeping spec compliance and code quality review gates enforced.

Two-stage quality gating blocks progress because each task must pass spec compliance review followed by code quality review before being considered complete. This gating logic ensures implementation plans meet requirements before tasks are accepted, reducing time-to-merge for independent development work.

team-dev reduces time-to-merge for implementation plans by running independent tasks in parallel through spawned implementer teammates while enforcing spec compliance and code quality through explicit review gates, using a lead orchestrator with staggered reviewer dispatch for high throughput.