| name | simplify |
| description | This skill should be used when the user wants to clean up code without changing behavior. Dispatches cleanup agents, runs tests to verify behavior is preserved, then runs code review. Triggers include "simplify this", "clean up this code", "refactor without changing behavior", "reduce complexity". |
Simplify
Orchestrate code cleanup through agent delegation. Preserve behavior. Don't change what the code does.
Act as orchestrator. Dispatch work to sub-agents via the Task tool. Do not edit code directly.
Input
Invoked as /simplify with optional arguments:
- No args: simplify files with uncommitted changes (
git status)
- File pattern: simplify files matching the pattern
- Notes path (
.lore/work/notes/*.md): resume a previous session
Process
Cleanup. Dispatch code-simplifier:code-simplifier via Task tool with the file list. If .lore/lore-agents.md exists, check for additional simplification agents in the Code Quality section and dispatch them too.
Test. Dispatch a testing agent via Task tool. Include which files changed and how to run the test suite. Expect pass/fail and notable findings. If tests fail, diagnose and route back to the cleanup agent for correction. After two failed attempts on the same issue, escalate to the user.
Review. Dispatch a review agent via Task tool. Include which files to review and the requirement that behavior must be preserved. Route findings back to the cleanup agent for correction. After two failed attempts, escalate to the user.
Output
Record what happened in a notes file at .lore/work/notes/simplify-<identifier>.md. Load ${CLAUDE_PLUGIN_ROOT}/shared/frontmatter-schema.md for the frontmatter fields. The body is freeform — capture which files were processed, what was simplified, any failures and how they were resolved.
Write the body in Markdown per the "Body Format" section of ${CLAUDE_PLUGIN_ROOT}/shared/frontmatter-schema.md. Reach for embedded inline HTML only when a visual carries meaning prose can't, such as a color-coded per-file outcome table.
Escalation
Two conditions require human intervention. Everything else is autonomous.
- Stuck loop: Two consecutive failed attempts on the same test or review failure.
- Behavior change detected: Tests fail in a way that suggests behavior changed rather than a cleanup regression.
