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 agentmods add skills/hajekim/agentic-design-patterns-extension/appendix-gui-agentsnpx skills add hajekim/agentic-design-patterns-extension --skill appendix-gui-agentsgit clone --depth 1 https://github.com/hajekim/agentic-design-patterns-extensionWrote 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/hajekim/agentic-design-patterns-extension/appendix-gui-agents)<a href="https://agentmods.dev/skills/hajekim/agentic-design-patterns-extension/appendix-gui-agents"><img src="https://agentmods.dev/badge/skills/hajekim/agentic-design-patterns-extension/appendix-gui-agents.svg" alt="Measured on agentmods" 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 | $0.00422 | $0.03495 |
| Opus 5 | $0.00211 | $0.01747 |
| Sonnet 5 | $0.00084 | $0.00699 |
| Haiku 4.5 | $0.00042 | $0.00349 |
Grade A, and why
appendix-gui-agents 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 5d 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.
This is a copy
100% identical to appendix-gui-agents — 3 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 340 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Appendix B - AI Agentic Interactions: From GUI to Real World
Overview
GUI and Real-World Interaction enables agents to transcend API-based automation and interact directly with graphical user interfaces and physical environments — just as humans do. This represents a fundamental shift: instead of requiring specialized API access for every system, agents can use the visual "front door" of any software, making them universally adaptable.
This pattern is powered by Agent-Computer Interfaces (ACIs) — the layer that allows AI to perceive visual elements (screenshots), reason about them (what is this button?), and act on them (click, type, scroll). Beyond digital interfaces, real-world agents extend to cameras, microphones, and physical sensor inputs.
Core Principle: Any software a human can use visually, an agent with computer use capabilities can operate — without requiring API access or custom integrations.
When This Skill Applies
Activate this pattern when:
- The target system has no API but has a visual interface (legacy software, web apps)
- Automating complex workflows that span multiple applications
- Building agents that must operate like a human user (form filling, navigation, data extraction from GUIs)
- Integrating with systems where API access is unavailable or too restrictive
- Building multimodal agents that respond to visual + voice + text inputs simultaneously
- Prototyping new workflows that will later be formalized with APIs
Rule of thumb: If a human can do it by looking at a screen and clicking, a GUI agent can automate it — but expect higher latency and error rates than API-based automation.
Agent-Computer Interface (ACI) Architecture
┌─────────────────────────────────────────────┐
│ ACI Processing Pipeline │
│ │
│ 1. Visual Perception │
│ └── Screenshot capture │
│ │
│ 2. GUI Element Recognition │
│ └── Identify buttons, fields, links │
│ │
│ 3. Contextual Interpretation │
│ └── LLM: "magnifying glass = search" │
│ │
│ 4. Dynamic Action & Response │
│ └── Click, type, scroll, drag │
│ │
│ 5. Feedback Monitoring │
│ └── New screenshot → verify action │
└─────────────────────────────────────────────┘
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.
- 5d ago First seen · 340 lines · 422 tokens per session scan A dced02785908
appendix-gui-agents is a skill published in the GitHub repository hajekim/agentic-design-patterns-extension (1 stars, last pushed 5mo ago), licensed MIT. It adds 422 tokens to every session and 3,495 once invoked, about $0.0021 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to appendix-gui-agents, differing in 3 lines, and is treated as a copy.
Other skills, from other repositories
browser-qa
Automated visual testing and browser interaction verification.
appendix-gui-agents
This skill should be used when the user wants to build "GUI agent", "computer use agent", "browser automation agent", "desktop automation", "visual agent", "screen interaction agent", "web automation agent", "agent computer interaction", "ACI agent", "click automation", "multimodal agent with vision", "Project Mariner…
e2e-testing
Playwright E2E testing patterns, Page Object Model, configuration, CI/CD integration, artifact management, and flaky test strategies.
prompt engineering
Use this skill when asked to create, refine, analyze, or optimize prompts for Large Language Models (LLMs). This skill ensures adherence to prompt engineering best practices and enforces a rigorous design workflow.
redteam
Expertise in offensive security research, vulnerability analysis, CMS-focused application testing, and red team operations.
redteam-cms
Focused methodology for authorized CMS fingerprinting, component inventory, misconfiguration review, and vulnerability validation.