What is context-engineer?

It’s a skill that analyzes and optimizes how tokens are allocated across system prompts, tools, and memory to improve efficiency.

Do I need external dependencies?

No. Token estimates and analyses run with Python 3 stdlib only.

How do I save and compare optimization progress?

Use --snapshot with analyze to save a state, then compare snapshots with the compare command to project token savings.

context-engineer

npx machina-cli add skill cacheforge-ai/cacheforge-skills/context-engineer --openclaw

Files (1)

SKILL.md

3.0 KB

When to use this skill

Use this skill when the user wants to:

Understand where their context window tokens are going
Analyze workspace files (SKILL.md, SOUL.md, MEMORY.md, etc.) for bloat
Audit tool definitions for redundancy and overhead
Get a comprehensive context efficiency report
Compare before/after snapshots to measure optimization progress
Optimize system prompts for token efficiency

Commands

# Analyze workspace context files — token counts, efficiency scores, recommendations
python3 skills/context-engineer/context.py analyze --workspace ~/.openclaw/workspace

# Analyze with a custom budget and save a snapshot for later comparison
python3 skills/context-engineer/context.py analyze --workspace ~/.openclaw/workspace --budget 128000 --snapshot before.json

# Audit tool definitions for overhead and overlap
python3 skills/context-engineer/context.py audit-tools --config ~/.openclaw/openclaw.json

# Generate a comprehensive context engineering report
python3 skills/context-engineer/context.py report --workspace ~/.openclaw/workspace --format terminal

# Compare two snapshots to see projected token savings
python3 skills/context-engineer/context.py compare --before before.json --after after.json

What It Analyzes

System prompt efficiency — Length, redundancy detection, compression potential
Tool definition overhead — Count tools, per-tool token cost, identify unused/overlapping
Memory file bloat — MEMORY.md size, stale entries, optimization suggestions
Skill overhead — Installed skills contributing to context, per-skill token cost
Context budget — What % of model context window is consumed by static content vs available for conversation

Options

--workspace PATH — Path to workspace directory (default: ~/.openclaw/workspace)
--config PATH — Path to OpenClaw config file (default: ~/.openclaw/openclaw.json)
--budget N — Context window token budget (default: 200000)
--snapshot FILE — Save analysis snapshot to FILE for later comparison
--format terminal — Output format (currently: terminal)

Notes

Token estimates are approximate (~4 characters per token). For precise counts, use a model-specific tokenizer.
No external dependencies required — runs with Python 3 stdlib only.
Built by Anvil AI — context engineering experts. https://labs.anvil-ai.io

Source

git clone https://github.com/cacheforge-ai/cacheforge-skills/blob/main/skills/context-engineer/SKILL.mdView on GitHub

Overview

This skill analyzes where tokens go and audits workspace files for bloat. It provides a comprehensive context efficiency report and token-cost breakdown. It also supports before/after snapshots to quantify improvements and guide optimization.

How This Skill Works

The tool parses system prompts, tool definitions, memory files, and installed skills to compute per-item token costs. It flags redundancy, overlap, and bloat, then suggests compression, removal, or reordering. Users can generate reports and save snapshots to track optimization progress over time.

When to Use It

Understand where context window tokens are going and visualize token flow
Analyze workspace files (SKILL.md, SOUL.md, MEMORY.md, etc.) for bloated content
Audit tool definitions for overhead and redundancy
Generate a comprehensive context efficiency report
Compare before/after snapshots to measure token-savings and optimization progress

Quick Start

Step 1: Analyze workspace: python3 skills/context-engineer/context.py analyze --workspace ~/.openclaw/workspace
Step 2: Audit tools (optional before deep dive): python3 skills/context-engineer/context.py audit-tools --config ~/.openclaw/openclaw.json
Step 3: Generate report: python3 skills/context-engineer/context.py report --workspace ~/.openclaw/workspace --format terminal

Best Practices

Run analyze with a consistent budget and save a snapshot to enable future comparisons
Regularly audit MEMORY.md and tool definitions to catch drift and bloat
Prioritize system prompt optimization before heavy tool tinkering for large gains
Use snapshots to quantify token savings and track progress over time
Document changes and compare successive reports to maintain a clear optimization trail

Example Use Cases

A team reduces static content in system prompts after a dedicated context audit, freeing significant tokens for conversation
Redundant tools are identified and removed, shrinking per-tool token costs and improving responsiveness
Memory file bloat is trimmed by pruning stale entries and compressing long-lived data
A comprehensive report highlights bottlenecks across skills and prompts, guiding targeted refactors
Snapshots before/after show measurable token savings and cleaner context budgets for agents

Frequently Asked Questions

Add this skill to your agents