When your Claude conversation gets bloated with token usage, this skill diagnoses the problem, shows you a dry-run of what it'll prune, then treats the session and spawns a watcher that auto-opens a fresh terminal window when you exit. You pick gentle, standard, or aggressive depending on how much history you want to keep. It's genuinely clever that it combines the treatment and reload in one command so you don't accidentally double-treat. The workflow is transparent: you see the token savings before committing, then just type /exit and you're back in business with a leaner context. Great for long debugging sessions that have accumulated too much noise.
npx -y skills add ruya-ai/cozempic --skill reload --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
sentry/dev
degausai/wonda
github/awesome-copilot
insforge/agent-skills
github/awesome-copilot
breaking-brake/cc-wf-studio