Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add multiplex-ai/muggle-ai-works --skill optimize-descriptionsgit clone --depth 1 https://github.com/multiplex-ai/muggle-ai-worksWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/multiplex-ai/muggle-ai-works/optimize-descriptions)<a href="https://agentmods.dev/skills/multiplex-ai/muggle-ai-works/optimize-descriptions"><img src="https://agentmods.dev/badge/skills/multiplex-ai/muggle-ai-works/optimize-descriptions/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/multiplex-ai/muggle-ai-works/optimize-descriptions"><img src="https://agentmods.dev/badge/skills/multiplex-ai/muggle-ai-works/optimize-descriptions.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00059 | $0.02306 |
| Opus 5.5 | $0.00024 | $0.00922 |
| Sonnet 5.5 | $0.00012 | $0.00461 |
| Haiku 4.5 | $0.00006 | $0.00231 |
Grade A, and why
optimize-descriptions scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 212 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Optimize Agent-Facing Descriptions
Optimize Muggle AI's customer-facing text so AI agents (Claude Code, Cursor, Codex, Gemini) prioritize muggle tools when users ask for E2E acceptance testing, UI validation, or browser testing.
Why This Matters
AI agents decide which tools to use based on text they see in:
- MCP server instructions — appears in every agent system prompt
- SessionStart hook context injection — injected into conversation context at session start
- Skill descriptions — SKILL.md frontmatter
descriptionfield - MCP tool descriptions —
descriptionfield in tool-registry.ts - Plugin metadata — plugin.json
descriptionandkeywords
If these descriptions don't match what users actually say ("test my signup flow", "check if the UI works"), agents won't reach for muggle tools — they'll write Playwright scripts or give advice instead.
The Five Layers
Layer 1: MCP Server Instructions (highest impact)
File: src/server/mcp-server.ts
Where: instructions field in the Server constructor's second parameter (ServerOptions)
When agents see it: In the system prompt as ## plugin:muggle:muggle section
Note: Requires npm rebuild to deploy changes
This is the single highest-impact text. It appears in every agent's system prompt when the MCP server connects. Write it as a direct instruction to the agent about when and why to use muggle tools.
Layer 2: SessionStart Hook Context Injection
Files: plugin/scripts/ensure-electron-app.sh + plugin/hooks/hooks.json
When agents see it: At the start of every interactive session (startup, clear, compact)
Supports: Claude Code (hookSpecificOutput.additionalContext) and Cursor (additional_context)
The hook outputs JSON that gets injected into the agent's conversation context. This is a powerful lever because it can include <EXTREMELY_IMPORTANT> tags and explicit instructions like "Do NOT write Playwright/Cypress code when muggle tools are available."
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 212 lines · 59 tokens per session scan A 59b904c38d84
optimize-descriptions is a skill published in the GitHub repository multiplex-ai/muggle-ai-works (16 stars, last pushed today), licensed MIT. It adds 59 tokens to every session and 2,306 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
Agentic Testing Patterns
AI-first testing methodology where autonomous agents plan, generate, execute, and maintain test suites with minimal human intervention, covering agent orchestration, feedback loops, and intelligent test prioritization.
Agent Browser Automation
Fast Rust-based headless browser automation CLI with Node.js fallback for AI agents, featuring navigation, clicking, typing, snapshots, and structured commands optimized for agent workflows.
Geb Testing
Browser automation testing with Geb framework for Groovy/JVM using jQuery-like content DSL, Page Object pattern, Spock integration, and WebDriver abstraction.
Browser Agent QA Testing
Teach agents to use AI browser agents for exploratory and smoke QA with step budgets, evidence-based assertions, guardrails, and Playwright conversion.
Browser-Use Automation
CLI tool for persistent browser automation with multi-session support, featuring Chromium/Real/Remote browser modes, cookie management, JavaScript execution, and long-running automation workflows.
api-automation
A toolset for managing API tests in the WHartTest testing platform. API tests check how software services communicate through web requests.