What happens when all tasks in a plan are finished?

Once all tasks are completed and verified, the skill automatically triggers the 'finishing-a-development-branch' sub-skill to run final tests and prepare the work for completion.

What is semantic agent selection?

Instead of just matching keywords, the skill analyzes the task's requirements (language, complexity, and type) to choose the best-suited agent, such as a Rust expert or a technical writer.

Is code review required during the execution process?

Yes, code review is mandatory after every batch of tasks. The skill will not proceed to the next batch until blocking issues are resolved and the human partner provides feedback.

How does this skill handle deviations from the original plan?

If an agent encounters a problem or believes a deviation is necessary, it reports a 'STOPPED' status. The skill then pauses and presents the user with the reasoning and options to approve, revise, or enforce the original plan.

Structured Plan Execution

Name: Structured Plan Execution
Author: tobyhede

bytobyhede

생산성 및 워크플로

Executes complex implementation plans in controlled batches with mandatory review checkpoints to ensure precision and architectural alignment.

소개

The Structured Plan Execution skill transforms static implementation plans into manageable workflows by processing tasks in logical batches. It ensures high-quality output by employing semantic agent selection—matching specific tasks to specialized agents for Rust, general coding, debugging, or documentation—rather than relying on simple keyword matching. By enforcing strict plan adherence and requiring human-in-the-loop checkpoints after every batch, it prevents unauthorized architectural drift and allows for critical review before proceeding, making it ideal for large-scale or complex development tasks.

주요 기능

Mandatory code review checkpoints between every execution batch
0 GitHub stars
Semantic agent selection based on task type and complexity
Automated verification and final development branch cleanup
Robust deviation handling with the 'STOPPED' status protocol
Controlled batch execution with a default three-task limit

사용 사례

Automating long-running development workflows that require frequent human verification and quality gates.
Managing complex feature implementations across polyglot codebases using specialized agents.
Executing multi-step architectural refactors that require strict adherence to a design document.

소개

주요 기능

Mandatory code review checkpoints between every execution batch
0 GitHub stars
Semantic agent selection based on task type and complexity
Automated verification and final development branch cleanup
Robust deviation handling with the 'STOPPED' status protocol
Controlled batch execution with a default three-task limit

사용 사례

Automating long-running development workflows that require frequent human verification and quality gates.
Managing complex feature implementations across polyglot codebases using specialized agents.
Executing multi-step architectural refactors that require strict adherence to a design document.