skill-condenser

Condenses oversized agent skill files (SKILL.md) so each invocation costs fewer tokens while preserving every behavior. Rewrites the frontmatter description for accurate triggering, tightens prose into tables, merges duplicate examples, and moves rarely needed detail to a marked appendix section at the end of the file. Use when a skill is bloated or monolithic, token cost per request is too high, the description misses or over-fires on triggers, or the user asks to optimize, tighten, slim, or refactor an existing skill. Not for writing skills from scratch, reviewing without editing, or splitting a skill across multiple files.

Version

v1.0.0

Size

9 KB

Updated

2026-10-06

Your client verifies the signature before reading. This explicit manual read is available on every plan. Free automatically loads five chosen skills; paid plans automatically load relevant skills across the library.