context-window-tracker
by Rian O'Leary
Visual context window monitoring and token breakdown for OpenClaw agents.
- Monitor real-time token usage with color-coded visual progress bars
- Identify which system files (like SOUL.md or TOOLS.md) consume the most space
- Predict session longevity with "turns remaining" estimates based on usage trends
Free
One-time purchase Β· Own forever
See it in action
π’ [ββββββββββββββββββββ] 52.4K / 128K tokens (41% used) ~42 turns left | Cache: 94% Detailed Breakdown: β’ System: 12K (AGENTS.md: 2K, MEMORY.md: 4K) β’ Conversation: 35.4K β’ Framework: 5K Trend: Growing at ~1.5K/turn. Plenty of room for the current refactor. See screenshots for better visual example β¬οΈβ¬οΈ
context-window-tracker
by Rian O'Leary
Visual context window monitoring and token breakdown for OpenClaw agents.
Free
One-time purchase Β· Own forever
β‘ Also available via Agensi MCP β your AI agent can load this skill on demand via MCP. Learn more β
Included in download
- Downloadable skill package
- 2 permissions declared
- Instant install
See it in action
π’ [ββββββββββββββββββββ] 52.4K / 128K tokens (41% used) ~42 turns left | Cache: 94% Detailed Breakdown: β’ System: 12K (AGENTS.md: 2K, MEMORY.md: 4K) β’ Conversation: 35.4K β’ Framework: 5K Trend: Growing at ~1.5K/turn. Plenty of room for the current refactor. See screenshots for better visual example β¬οΈβ¬οΈ
Screenshots
About This Skill
What it does
The Context Window Tracker provides real-time visibility into your AI agent's memory and token usage. It eliminates the guesswork of "will this next prompt crash my session?" by providing a visual, glanceable report of exactly how much context remains and what is consuming your token budget.
- Ideal use for developing outside of openclaw's TUI where there is no built in visual tracker (slack, discord, Whatsapp etc)
Why use this skill
Unlike basic token counters, this skill performs a deep forensic analysis of your session. It breaks down usage into specific components: system prompts (including individual file contributions like IDENTITY.md or MEMORY.md), conversation history, and framework overhead. It even tracks "thinking" token trends and calculates an estimated "turns remaining" based on your current conversation velocity.
Supported tools
- OpenClaw Framework: Deep integration with session stores and .jsonl transcripts.
- Provider Agnostic: Works across different LLM providers by reading local session metadata.
- Automatic Monitoring: Features a built-in message counter to auto-report status every 10 messages.
The Output
You get a clean, CLI-style visual report. In Compact Mode, it's a single line with a color-coded progress bar. In Detailed Mode, you receive a full hierarchical breakdown of file-by-file token costs and growth trends, allowing you to identify exactly which files or interactions are bloating your context.
π Learn more: Best DevOps & Deployment Skills for Claude Code β
Use Cases
- Monitor real-time token usage with color-coded visual progress bars
- Identify which system files (like SOUL.md or TOOLS.md) consume the most space
- Predict session longevity with "turns remaining" estimates based on usage trends
- Automate usage reporting every 10 messages to prevent session overflows
- Analyze cache hit rates and thinking token overhead for cost optimization
Known Limitations
Cant be used on agents with web UI's such as Chatgpt.com or Claude.ai. Best used in frameworks like Clawhub, Claude code etc
How to Install
unzip context-window-tracker.zip -d ~/.claude/skills/Reviews
No reviews yet β be the first to share your experience.
Only users who have downloaded or purchased this skill can leave a review.
No reviews yet β be the first to share your experience.
Only users who have downloaded or purchased this skill can leave a review.
Security Scanned
Passed automated security review
Permissions
This skill works best in frameworks such as openclaw.
Frequently Asked Questions
Learn More About AI Agent Skills
Similar Skills
env-doctor
Diagnoses why your project will not start. Checks runtime versions, dependencies, environment variables, databases, ports, and build artifacts systematically.
git-commit-writer
Writes conventional commit messages by analyzing your staged git changes. Detects commit type, scope, and breaking changes automatically.
ai-productivity
High-speed intake for shaping vague prompts, triaging complex tasks, and compressing context for efficient execution.
code-reviewer
Reviews your code for bugs, security vulnerabilities, logic errors, performance issues, and style violations. Organizes findings by severity and suggests fixes with code examples.