Agent skill

diagnose

Analyze Claude Code session bloat — shows token count, context usage %, and bloat breakdown. Use when the user asks about session size, context usage, or when you notice the context window is getting full.

View SKILL.md on GitHub Repository
Stars 222
Forks 12

Install this agent skill to your Project

npx add-skill https://github.com/Ruya-AI/cozempic/tree/main/plugin/skills/diagnose

SKILL.md

Run a diagnosis on the current session:

bash
cozempic current --diagnose

The output includes:

  • Weight: total session size in bytes and message count
  • Tokens: exact token count (from usage data) or heuristic estimate
  • Context bar: visual bar showing % of the detected context window (200K or 1M)
  • Vital signs: progress ticks, file history snapshots, system reminders, thinking content, signatures, tool results
  • Message type breakdown: bytes per message type
  • Top 10 largest messages: biggest bloat contributors
  • Estimated savings by prescription: what gentle/standard/aggressive would save

Always surface the token count and context % to the user. If context is above 60%, suggest running /cozempic:treat with a prescription recommendation:

  • Under 5MB: gentle
  • 5-20MB: standard
  • Over 20MB: aggressive

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

Ruya-AI/cozempic

guard

Protect Claude Code sessions from context overflow by running a background daemon that monitors session size and auto-prunes before compaction hits. Use when the user says "guard", "protect session", "context getting long", "prevent compaction", "session management", or is running agent teams that need continuous context protection.

222 12
Explore
Ruya-AI/cozempic

reload

Treat the current session and auto-resume in a new terminal window.

222 12
Explore
Ruya-AI/cozempic

treat

Prune bloated session with a prescription. Removes progress ticks, stale reads, duplicate content, and more.

222 12
Explore
Ruya-AI/cozempic

doctor

Run health checks on Claude Code configuration and sessions. Use when troubleshooting Claude Code issues.

222 12
Explore
davila7/claude-code-templates

verl-rl-training

Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.

23,776 2,298
Explore
davila7/claude-code-templates

openrlhf-training

High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.

23,776 2,298
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results