Skip to main content
skill CX Codex

context-optimizer — Claude Code skill: what it does, how to install

context-optimizer controls token consumption and frees context in Claude Code. Use it when agent sessions slow down, forget instructions, or approach budget limits.

MIT ★ 2,966 updated 2026-09-29

Install into Claude Code · Plugin marketplace All methods and platforms ↓
/plugin marketplace add rohitg00/pro-workflow
/plugin install pro-workflow@pro-workflow
Open on GitHub

What context-optimizer does

The skill sets clear context management protocols for Claude Code sessions. It guides users through diagnosing memory usage with /context, running /compact when utilization passes 70%, and switching to a fresh /resume session once unrelated tasks begin.

It outlines strict rules for output length and external tools. The instructions demand keeping fewer than 10 MCP servers and under 80 tools active at once, because every enabled MCP adds overhead to each request. Heavy outputs from test runners, logs, and docs are pushed to isolated subagents to preserve the main session.

Users can configure proactive compaction via the CLAUDE_AUTOCOMPACT_PCT_OVERRIDE environment variable set to 50. Without trimming root CLAUDE.md files below 150 lines, repetitive tool failures and generic responses appear much sooner.

What it helps with

  • trim token consumption across active Claude Code sessions
  • trigger context compaction before memory usage passes 70%
  • offload heavy test and log outputs to subagents
  • audit active MCP tools to reduce per-request overhead

Example prompts for the agent

  • Check our current context usage and clean up token overhead
  • Audit active MCP servers and reduce prompt bloat
  • Delegate test runner execution to a subagent to save main context

How to install context-optimizer

1Plugin marketplace

/plugin marketplace add rohitg00/pro-workflow
/plugin install pro-workflow@pro-workflow

This installs the whole pro-workflow plugin with all its skills

2Skills CLI (Vercel)

npx skills add rohitg00/pro-workflow --skill context-optimizer -a claude-code

3Manually: put the files in place

~/.claude/skills/context-optimizer/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/context-optimizer/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about context-optimizer

How to install context-optimizer in Claude Code?

First, add the repository via /plugin marketplace add rohitg00/pro-workflow. Then run /plugin install pro-workflow@pro-workflow to finish installation.

What does context-optimizer do?

It provides structured token budget targets, diagnostic thresholds for /context, and actionable rules to prevent degraded agent responses.

How do you enable automatic context compaction?

Set the CLAUDE_AUTOCOMPACT_PCT_OVERRIDE environment variable to 50 in your configuration. This triggers proactive compaction at half capacity.

Source text

SKILL.md · MIT
---
name: context-optimizer
description: Optimize token usage and context management. Use when sessions feel slow, context is degraded, or you're running out of budget.
user-invocable: true
---

# Context Optimizer

Manage your context window and token budget effectively.

## Quick Diagnosis

1. Run `/context` to check current usage
2. If > 70% → compact now before it degrades
3. If > 90% → you're in the "dumb zone", compact immediately

## Optimization Strategies

### Immediate

| Action | Saves | When |
|--------|-------|------|
| `/compact` | 30-50% context | At task boundaries |
| Disable unused MCPs | ~5% per MCP | When switching domains |
| Use subagents for exploration | Keeps main context clean | Heavy search/read tasks |
| Fresh session via `/resume` | 100% reset | When starting unrelated work |

### Configuration

Set proactive auto-compaction:
```json
{
  "env": {
    "CLAUDE_AUTOCOMPACT_PCT_OVERRIDE": "50"
  }
}
```

### MCP Audit

Keep <10 MCPs enabled, <80 tools total. Each MCP adds overhead to every request.

```bash
/mcp          # List active servers
# Disable what you're not using
```

### Prompt Engineering for Token Efficiency

- Scope your prompts: "In src/auth/, fix the login bug"
- Provide constraints: "Don't modify the middleware"
- Give acceptance criteria: "Should return 429 after 5 attempts"
- Avoid vague prompts: "Fix the code" (forces Claude to read everything)

### Subagent Delegation

Heavy operations that generate lots of output should go to subagents:

- Test suite output → subagent
- Large file exploration → subagent
- Documentation generation → subagent
- Log analysis → subagent

The main session stays clean while subagents handle the volume.

## Context Budget Planning

| Phase | Target Usage | Action If Over |
|-------|-------------|----------------|
| Planning | < 20% | Keep plans concise |
| Implementation | < 60% | Compact between files |
| Testing | < 80% | Delegate to subagent |
| Review | < 90% | Start fresh session |

## Token Efficiency

### Output Reduction (40-60% savings)
- No sycophantic openers ("Sure!", "Great question!")
- No closing fluff ("Let me know if you need anything!")
- No prompt restatement before answering
- Code first, explanation only if non-obvious
- Structured output (tables, bullets) over prose
- ASCII only: -- not em dashes, straight quotes not smart quotes

### Behavioral Efficiency
- One-pass coding: complete solution, test once, stop if green
- Read before write: never modify unread files
- No re-reads: don't re-read unchanged files
- Tool-call budgets: 20 (quick fix) to 80 (large feature)
- Never iterate more than twice on the same failure

### Task Profiles
Switch response style based on context:
- **Coding**: code first, minimal explanation, simplest solution
- **Agent/Pipeline**: structured output only, no prose, parseable
- **Analysis**: finding first, tables over paragraphs, sourced numbers

## CLAUDE.md Optimization

- Root CLAUDE.md: < 60 lines ideal, < 150 max
- Move package-specific info to package-level CLAUDE.md
- Move personal preferences to CLAUDE.local.md
- Remove obvious or rapidly-changing information

## When Context Is Degraded

Signs:
- Claude repeats itself or forgets earlier context
- Responses become generic or lose project-specific knowledge
- Tool calls start failing for reasons that worked earlier

Fix:
1. Manual `/compact`
2. If still bad: new session with `/resume`
3. For recurring issues: reduce CLAUDE.md size, disable MCPs

© Rohit Ghumare · MIT · rohitg00/pro-workflow

Similar skills

All skills for Agent setup: memory, context, orchestration →

From the same set83

The whole set →
reviewer reviewer — a subagent that reviews code line by line, reading actual files and citing every finding with exact quotes. Use before committing, during PR review, or after major changes. CX Codex ★ 2,966repo stars skill-conventions skill-conventions defines standards for workflow skills: separates human invocation from auto-triggers, governs state writes, and keeps the skill router in sync. CX Codex +1 ★ 2,966repo stars plan-interrogate The plan-interrogate skill probes a plan before implementation by mapping its decisions and asking about them one at a time. Use it to clarify unresolved design choices. CX Codex +1 ★ 2,966repo stars bug-capture bug-capture is a Claude Code skill that turns reported defects into GitHub issues using the project’s own terms. It activates after a conversational bug report, a QA pass, or a request to file an issue. CX Codex +1 ★ 2,966repo stars deslop deslop trims unnecessary code from the current branch diff and checks SKILL.md files. Run it before committing or reviewing a skill for release. CX Codex +1 ★ 2,966repo stars wiki-builder wiki-builder is a Claude Code skill that creates a structured knowledge base from Markdown pages and an FTS5 index in SQLite. It persists between sessions and is queryable by the agent. CX Codex +1 ★ 2,966repo stars token-efficiency token-efficiency cuts conversational fluff, caps tool executions per task, and stops Claude Code from repeatedly editing working implementations. CX Codex +1 ★ 2,966repo stars wiki wiki is a Claude Code command that builds and maintains persistent research wikis. It stores markdown files, indexes them in SQLite FTS5, and retrieves context via BM25. CX Codex +1 ★ 2,966repo stars

Project sponsors

Companies that keep this analytics open