cavecrew

Context exhaustion during codebase analysis is managed by delegating tasks to specialized subagents. This approach compresses subagent output by 60% compared to vanilla prose, significantly reducing context window consumption while locating code paths.

Code editing delegation to a subagent is recommended when you need precise edits in a large codebase and want to optimize context. A decision matrix triggers specialized agents like the builder for edits and the investigator for code path location.

Subagent delegation for code analysis works by triggering specialized agents based on a decision matrix. Agents like the investigator rapidly locate symbols and code paths, compressing output by 60% to prevent context exhaustion during large codebase analysis.

Yes, specific subagents can be used for different code editing tasks. The system provides specialized agents such as an investigator to locate symbols, a builder to make precise edits, and a reviewer, ensuring efficient task specialization.

The best way to locate a specific symbol in a large codebase is to invoke a specialized investigator subagent. This delegates the code analysis task, compressing output by 60% to preserve context while rapidly finding the relevant code paths.

Delegating code analysis reduces context exhaustion because subagent output is compressed by 60% compared to vanilla prose. By using a decision matrix to trigger specialized agents, context size is optimized during rapid code segment location and editing.