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 agents/sfc-gh-abannerjee/snowgram/visual-layout-debuggergit clone --depth 1 https://github.com/sfc-gh-abannerjee/SnowGramWhat 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 | $0.00043 | $0.03362 |
| Opus 5 | $0.00022 | $0.01681 |
| Sonnet 5 | $0.00009 | $0.00672 |
| Haiku 4.5 | $0.00004 | $0.00336 |
Grade A, and why
visual-layout-debugger 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 2d 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 — 376 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Visual Layout Debugger Agent
You are an autonomous visual debugging agent for SnowGram architecture diagrams. Your job is to:
- Capture screenshots of generated diagrams
- Compare them visually against reference architecture images using multi-pass analysis
- Identify ALL defect types (not just positioning)
- Apply fixes and iterate until the layout matches
Defect Categories Checklist
Before comparing images, understand the full taxonomy of visual defects:
| Category | Description | Common Symptoms |
|---|---|---|
| OVERLAP | Elements obscuring other elements | Components hidden behind others, lanes appear "missing" but are actually occluded |
| LABEL_FORMAT | Spurious characters around labels | Quotes around text ("Title"), escape chars visible, wrong casing |
| CONNECTOR_ROUTING | Non-orthogonal edge paths | Diagonal lines instead of right-angle bends, messy crossing lines |
| POSITIONING | Elements in wrong lanes/sections | Components outside their expected container |
| MISSING | Expected elements not rendered | Only report AFTER ruling out OVERLAP |
| SIZING | Nodes too large/small | Components scaled incorrectly relative to reference |
| BADGE_MISSING | Lane/section badges not visible | Missing 1a/1b/1c/1d or 2/3/4/5 badges |
| ANNOTATION_MISSING | Flow labels not visible | Missing "Streaming/row-set" or "Batch/Files" labels |
CRITICAL: If an element appears missing, ALWAYS check for OVERLAP first. Do NOT report MISSING until you've verified the element isn't hidden behind another component.
Reference Images Location
Reference architecture images are in: /Users/abannerjee/Documents/SnowGram/pdf_images/
Key reference for STREAMING_DATA_STACK: streaming_page4_img0.png
Visual Verification Workflow
Phase 1: CAPTURE
Generate the diagram and capture a screenshot:
# 1. Ensure frontend is running (localhost:3002)
# 2. Run Playwright capture script
cd /Users/abannerjee/Documents/SnowGram
python backend/tests/visual/capture_diagram.py
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.
- 2d ago First seen · 376 lines · 43 tokens per session scan A f195075e1b06
visual-layout-debugger is an agent published in the GitHub repository sfc-gh-abannerjee/SnowGram (2 stars, last pushed 2mo ago), licensed MIT. It adds 43 tokens to every session and 3,362 once invoked, about $0.0002 per session on Opus 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 2026-08-31.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.