Markdown Token Optimizer
Shrink long Markdown skill and doc files before they fill the agent context window during Azure-skills workflows.
Overview
markdown-token-optimizer is an agent skill most often used in Build (also Operate, Ship) that trims or restructures Markdown so coding agents spend fewer tokens on the same procedural content.
Install
npx skills add https://github.com/microsoft/azure-skills --skill markdown-token-optimizerWhat is this skill?
- Reduces Markdown bulk while preserving structure agents need to follow procedures
- Pairs with the Microsoft azure-skills repo pattern of large SKILL.md and reference trees
- Improves cost and latency on repeated doc loads in Claude Code, Cursor, and Codex
- Useful before committing or publishing verbose documentation packs
- Complements human-written docs rather than replacing authoring standards
Adoption & trust: 1k installs on skills.sh; 1.2k GitHub stars.
What problem does it solve?
Your SKILL.md and reference Markdown are so long that agents drop steps, hallucinate gaps, or burn most of the context window before touching code.
Who is it for?
Builders maintaining large azure-skills-style Markdown trees who hit context limits in daily agent sessions.
Skip if: First-time README authoring or teams that need full GFM tutorials—use a dedicated Markdown documentation skill instead.
When should I use this skill?
Agent Markdown files routinely exceed comfortable context share or duplicate content across SKILL.md and references.
What do I get? / Deliverables
You get leaner agent-ready Markdown that preserves navigable structure and instructions, leaving more context for implementation and review.
- Token-reduced Markdown draft
- Section map or merge notes for what was condensed
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Token-efficient Markdown is authored and maintained alongside agent skills in the Build phase, where SKILL.md and reference trees dominate context. Fits agent-tooling because the skill targets Markdown consumed by coding agents, not end-user product UI.
Where it fits
Collapse repetitive sections in a SKILL.md before wiring a new Azure validate skill.
Trim reference Markdown after importing a long Microsoft template tree.
Shorten release notes copies agents load during final checklist runs.
Refresh ops runbooks so incident skills fit beside logs in one session.
How it compares
Context-budget optimizer for agent Markdown, not a general Markdown style or GFM reference skill.
Common Questions / FAQ
Who is markdown-token-optimizer for?
Solo and indie developers using Claude Code, Cursor, or Codex with Microsoft azure-skills or similarly large Markdown skill packages who need smaller agent payloads.
When should I use markdown-token-optimizer?
During Build when refining SKILL.md and references; before Ship when release docs swell; in Operate when runbooks grow—whenever agent context is tight but you must keep procedural detail.
Is markdown-token-optimizer safe to install?
Review the Security Audits panel on this Prism page and inspect the skill source in the microsoft/azure-skills repository before enabling it in automated agent runs.
Workflow Chain
Then invoke: markdown documentation
SKILL.md
READMESKILL.md - Markdown Token Optimizer
# markdown-token-optimizer Installs: 1012 Source: microsoft/azure-skills { "id": "microsoft/azure-skills/markdown-token-optimizer", "url": "https://www.skills.sh/microsoft/azure-skills/markdown-token-optimizer", "name": "markdown-token-optimizer", "slug": "markdown-token-optimizer", "source": "microsoft/azure-skills", "installs": 1012, "installUrl": "https://github.com/microsoft/azure-skills", "sourceType": "github" }