dispatching-parallel-agents

What problem does it solve? Investigating multiple unrelated failures sequentially wastes time and bloats your working context. This Skill provides a structured pattern for splitting independent problems across concurrent agents, each with precisely crafted instructions and no inherited session history. ## Core Features & Use Cases - Independence Decision Framework: A decision flowchart and criteria to determine whether failures are independent enough for parallel dispatch or require sequential investigation. - Agent Prompt Structure: A proven template for writing focused, self-contained agent tasks with specific scope, constraints, and expected output format. - Common Mistakes Guide: Concrete examples contrasting vague prompts ("fix all the tests") with effective ones (specific file, error messages, constraints). - Use Case: After a major refactoring leaves 6 test failures across 3 files, dispatch one agent per test file concurrently, then review summaries, check for conflicts, and run the full suite to integrate all fixes. ## Quick Start Ask the AI to dispatch one parallel agent per failing test file, giving each agent the specific test names, error messages, and a constraint to not modify unrelated code.