task-scaler_skill

Overview

This skill evaluates task complexity and assigns a scale classification to optimize agent usage, workflow design, and resource allocation. It produces a numeric complexity score, a factor breakdown, and a workflow recommendation to guide execution planning. Results help teams decide direct execution, number of agents, and testing or integration needs.

How this skill works

The skill parses the task description for keywords and patterns to produce an initial scale guess (trivial, small, medium, large). It then measures concrete factors—files affected, dependency depth, test requirements, user interaction, integrations, and database impact—and computes a weighted complexity score. Contextual rules can scale the result up or down based on ambiguity, existing patterns, or security risk, and the skill returns factor details and actionable workflow recommendations.

When to use it

• Estimating work before assigning agents or creating tickets
• Routing tasks to direct execution vs. multi-agent workflows
• Planning testing and integration effort for new features
• Deciding when to require RAI or specialized review
• Batching similar small fixes to reduce fragmentation

Best practices

• Be conservative—prefer scaling up if requirements are unclear
• Use repository history or patterns to scale down predictable work
• Prefer minimal agent usage; batch similar tasks for one agent
• Explicitly flag security or schema changes to increase scale
• Record decisions and iteration estimates for continuous calibration

Example use cases

• A one-line typo or formatting fix classified as trivial for direct execution
• A single-file bug fix evaluated as small with deliverable review recommended
• A feature touching multiple modules scored medium and flagged for tests and RAI
• A new microservice or architecture change classified as large with full-workflow orchestration
• Batching three related refactors into one small/medium job to avoid agent overuse