What problem does it solve?
When exploring or understanding code you did not write, agents typically have to guess intent from code structure, which is often inaccurate and time-consuming. This skill eliminates that guesswork by directing agents to use Entire's recorded checkpoint history, which captures the original prompts, transcripts, and decisions behind every code change.
Core Features & Use Cases
- Sub-skill routing: Automatically delegates to specialized Entire skills for targeted tasks like searching past work, explaining code provenance, or reviewing branches with intent context.
- General exploration workflow: Provides a step-by-step process for understanding unfamiliar modules or codebases when no specialized sub-skill fits, using checkpoint data to back all findings.
- Transparent reporting: Enforces clear labeling of all insights as either checkpoint-backed recorded intent or code-inferred observations to avoid misrepresentation.
- Use Case: If you need to refactor an unfamiliar authentication module, use this skill to pull the original developer's intent from checkpoints instead of guessing why the code was structured that way.
Quick Start
Use the using-entire skill to explain the original intent behind the src/auth.ts file and clearly distinguish between checkpoint-backed context and observations inferred from the code structure.