Agent skill

Worldlabs Topology Viz

by automateyournetwork in automateyournetwork/netclaw

Turn a real network topology into a free, no-cost themed prompt preview, and optionally (after explicit confirmation) generate an explorable, AI-generated 3D 'world' via World Labs Marble — a…

Apache-2.0Auto-check: notesProduct & Project Management

Install Worldlabs Topology Viz

skills CLI
$ npx skills add automateyournetwork/netclaw --skill worldlabs-topology-viz -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install automateyournetwork/netclaw worldlabs-topology-viz --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/automateyournetwork/netclaw.git skills-src && mkdir -p .claude/skills && cp -r skills-src/workspace/skills/worldlabs-topology-viz .claude/skills/worldlabs-topology-viz && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
worldlabs-topology-viz
GitHub stars
676
Token cost
~2.9k tokens
SKILL.md length
1,321 words
Files
3
Skills in repo
120
Repo updated
First seen
Licence
Apache-2.0

At a glance

Turn a real network topology into a free, no-cost themed prompt preview, and optionally (after explicit confirmation) generate an explorable, AI-generated 3D 'world' via World Labs Marble — a…

  • Works in 2 steps: Preview — free, instant, no World Labs… → Generate — spends real World Labs…
  • The operator asks for a fantastical
  • SKILL.md covers Overview, Prerequisites, Workflow: Free Preview (User… and Workflow: Confirm and Generate…, plus 3 more sections
  • Runs Python scripts from its folder; needs WLT_API_KEY

What it does

Worldlabs Topology Viz is an agent skill from automateyournetwork/netclaw. Turn a real network topology into a free, no-cost themed prompt preview, and optionally (after explicit confirmation) generate an explorable, AI-generated 3D 'world' via World Labs Marble — a decorative companion visualization, never a substitute for the accurate topology diagram it is derived from. Use when the operator asks for a fantastical, explorable, or AI-generated 3D world/scene of a network topology.

Its SKILL.md is about 2.9k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `fantastical_prompt_builder.py` and `topology_model.py`).

It sits in Product & Project Management. It works with Model Context Protocol. The repository describes itself as: An AI agent that claws through your network. The licence is Apache-2.0.

When your agent uses it

  • The operator asks for a fantastical
  • AI-generated 3D world/scene of a network topology

Example prompts

  • “/worldlabs-topology-viz”

Requirements

  • Python 3
  • A credential in WLT_API_KEY

Workflow steps

2 steps, taken from the first numbered list in SKILL.md.

  1. Preview — free, instant, no World Labs call, repeatable with a different theme.
  2. Generate — spends real World Labs credits (~5 minutes per world), and only runs after an

What it can do on your machine

Read from SKILL.md and the folder at commit aa90e7d. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships script files (Python), which the agent can run.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names these keys or tokens, usually read from environment variables:

    • WLT_API_KEY

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Worldlabs Topology Viz loads about 2.9k tokens when it runs. Until then it costs about 109 tokens; SKILL.md has 1,321 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~109
When it runs · the whole SKILL.md, loaded when a task matches
~2.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

  • NoteMentions a .env fileSKILL.md:37
    - **`WLT_API_KEY`** in `.env` — your World Labs API key (`platform.worldlabs.ai/api-keys`), with a

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from automateyournetwork/netclaw at commit aa90e7d, republished under its Apache-2.0 licence (© automateyournetwork). 1,321 words, ~2,862 tokens.

Download SKILL.mdSave it as .claude/skills/worldlabs-topology-viz/SKILL.md (or your agent's skills folder). This skill also uses 2 other files; get the full folder from GitHub.
name
worldlabs-topology-viz
description
Turn a real network topology into a free, no-cost themed prompt preview, and optionally (after explicit confirmation) generate an explorable, AI-generated 3D 'world' via World Labs Marble — a decorative companion visualization, never a substitute for the accurate topology diagram it is derived from. Use when the operator asks for a fantastical, explorable, or AI-generated 3D world/scene of a network topology.
license
Apache-2.0
user-invocable
true

World Labs Fantastical Topology Visualization Skill

Version: 1.0.0 Feature: 122-worldlabs-topology-viz Status: Active

Overview

Produces an AI-augmented, explorable 3D "world" visualization of a real network topology, built on top of the existing spec 121 topology-diagram-mcp pipeline. World Labs' Marble API is a generative world model — it does not accept structured/graph input and has no mechanism for precise node placement, so this is explicitly a decorative, companion visualization, never a replacement for the accurate topology diagram. The real, structurally-correct diagram (produced by the existing, unmodified topology-diagram-mcp/render_structural tool) remains the authoritative artifact and is always referenced alongside any result this skill produces.

Two distinct modes:

  1. Preview — free, instant, no World Labs call, repeatable with a different theme.
  2. Generate — spends real World Labs credits (~5 minutes per world), and only runs after an explicit, separate user confirmation, enforced both conversationally and by the worldlabs-marble-mcp tool itself (it will not proceed without user_confirmed=true).

Prerequisites

  • WLT_API_KEY in .env — your World Labs API key (platform.worldlabs.ai/api-keys), with a funded account (platform.worldlabs.ai/billing). Required only for the generate mode — preview needs no credential at all.
  • The worldlabs-marble-mcp server (mcp-servers/worldlabs-marble-mcp/), registered in config/openclaw.json.
  • The existing topology-diagram-mcp server (spec 121), already registered — reused unmodified.
Topology sources

Any topology snapshot already normalized into the devices/links shape topology-diagram-mcp accepts (CML, pyATS, or any of NetClaw's other topology-of-record integrations, or a freeform description).

Workflow: Free Preview (User Story 1)

Zero cost, instant, repeatable with a different theme. No call to worldlabs-marble-mcp is ever made in this workflow.

  1. Obtain the topology snapshot (devices/links) from whatever source the operator named (or a freeform description), normalized into the shape topology-diagram-mcp/render_structural already accepts: devices: [{"hostname","role","state"}], links: [{"a","b","label"}].
  2. Call topology-diagram-mcp/render_structural(snapshot_id, devices, links) — reused unmodified. This tool has two existing failure modes that MUST both be surfaced clearly, distinctly, and before anything else in this workflow runs (FR-012 and its Edge Cases correction):
    • Empty topology (devices is empty): the tool raises "devices list is empty — nothing to render". Report exactly that — this topology has nothing to preview. Do not attempt to fabricate a prompt or a result.
    • Too many devices (over the working-resolution density ceiling, 60 devices): the tool raises a message naming the ceiling. Report that specific reason (e.g. "this topology has N devices, exceeding the 60-device legibility ceiling — narrow the scope and try again"), not a generic failure.
    • A single device with zero links is NOT a failure — render_structural already supports it (a one-node layout), and this workflow proceeds normally.
  3. On success, build a topology_model.TopologySnapshot from the same devices/links data (via topology_model.Device/Link), and call fantastical_prompt_builder.build_prompt(snapshot, theme=<operator-specified theme, or None for the default>).
  4. Present the result to the operator:
    • The reference diagram (render_structural's image_base64, or a note of where it's shown).
    • The composed text prompt from step 3.
    • topology_model.DECORATIVE_LABEL, verbatim, every single time (FR-009) — this is a preview of what could be generated, not a claim about what a generated world will precisely look like (Marble is non-deterministic).
    • An explicit statement that no credits were spent and no call was made to World Labs.

Workflow: Confirm and Generate (User Story 2)

This is the one part of this skill that spends real money. Only run this after Workflow: Free Preview has already produced a reference diagram and prompt for the same topology snapshot.

  1. State plainly, as its own message, separate from anything else: "Generating this world will spend World Labs credits and typically takes about 5 minutes. Proceed?" Do not combine this with the preview step's output, and do not proceed without an explicit affirmative reply (FR-004). This conversational check is the first of two independent layers — see step 2.
  2. Only once the operator has explicitly confirmed, call worldlabs-marble-mcp/generate_world(text_prompt=<from the preview's composed prompt>, display_name=<derived from the snapshot's identity, max 64 chars>, user_confirmed=true). Do NOT pass image_base64 (research.md R9/R10, corrected 2026-09-03 after live evidence from six real production generations): attaching the reference diagram as an image gets pasted flat and unchanged into the generated scene instead of being used as structural guidance, and is also measurably less reliable (3 of 4 image-bearing attempts failed; 4 of 4 text-only attempts succeeded on the first try). fantastical_prompt_builder.build_prompt already describes every real device and every real link individually in the text — that is the thing that actually carries real data into the result, not the image. The user_confirmed=true argument is the second, code-level layer (FR-016) — generate_world itself refuses to make any outbound call without it, so this is not optional plumbing to skip.
  3. Immediately after that call returns (success, failure, or a confirmation_required rejection — which should never happen if step 1 was actually followed, but is handled the same way if it does), record a GAIT entry via gait_record_turn (Constitution Principle IV, FR-015): user_text names the topology snapshot's identity/theme and that the operator confirmed; assistant_text names the operation_id and, once known, the world_id/world_marble_url/cost.total_credits or the failure category; artifacts is empty or references the reference diagram's identity only — never the API key, never raw device data beyond hostname/role identity. This step is required for every confirmed attempt, not optional, and is distinct from any end-of-session GAIT summary.
  4. Report the result:
    • Success: the operation_id (tell the operator to note it — nothing server-side tracks it, Clarifications Q1), and that status can be checked later.
    • confirmation_required: this means step 1/2 were skipped — report it and restart from step 1, do not retry the call with the flag silently added.
    • authentication_failure: tell the operator to check WLT_API_KEY — never repeat the key value itself.
    • insufficient_credits: pass through World Labs' own message (it already names the fix — add credits or enable auto-refill).
    • rate_limited: tell the operator to wait and retry later — do not resubmit automatically.
    • generic_failure: pass through the provider's message.
  5. To check status later, call worldlabs-marble-mcp/check_generation_status(operation_id).
    • done: false — still in progress; report that plainly.
    • done: true with a response — completed. Report world_marble_url (the viewer link), assets, and cost.total_credits, plus topology_model.DECORATIVE_LABEL verbatim, every time (FR-009) — a completed generation is exactly the case someone could mistake for an accurate diagram, so this label matters most here.
    • done: true with an error — the generation itself failed after starting; report the error plainly, and do not retry automatically.
    • not_found_or_expired — the operation record itself expired (they carry a roughly one-hour expires_at). If an earlier poll's metadata included a world_id, fall back to worldlabs-marble-mcp/get_world(world_id) instead of reporting a hard failure — the world itself is not necessarily gone just because the operation record is. Include DECORATIVE_LABEL here too if get_world succeeds.
Show full SKILL.md (256 more words)Show less

Workflow: Direct Prompt Generation (bring your own prompt)

For when the operator has already written a complete, ready-to-use prompt themselves (e.g. iterating creatively on wording) and wants it sent straight to Marble — no topology fetch, no render_structural call, no fantastical_prompt_builder, no other skill or data source involved at all. The operator's own words are the entire text_prompt, verbatim.

  1. Confirm credits/time exactly as in step 1 of Confirm and Generate above — this requirement does not change just because the prompt is hand-written.
  2. Call worldlabs-marble-mcp/generate_world(text_prompt=<the operator's prompt, verbatim>, display_name=<a short label the operator gives, or a reasonable default>, user_confirmed=true) — no image_base64, same reasoning as above.
  3. Record the GAIT entry and report the result exactly as steps 3-5 of Confirm and Generate above — those steps do not depend on where the prompt came from. DECORATIVE_LABEL still applies: a hand-written prompt with no topology data behind it at all is not a diagram substitute either.

Natural Language Commands

Free preview (no cost)
"Give me a fantastical world preview of the CML lab topology, floating-islands theme"
"Preview what this topology would look like as an underwater city, using the current pyATS testbed"
"Show me a fantastical preview of this topology: a router r1 connected to a switch sw1"
Generate (spends credits — requires confirmation)
"Yes, generate it"
"Go ahead and generate that world"
Direct prompt (bring your own prompt, spends credits — requires confirmation)
"<a full hand-written scene description>" send this straight to Marble and send me the link
"<a full hand-written scene description>" generate that, no need to pull any topology for this one
Check status
"Is the world done yet?"
"Check the status of that generation"

How this differs from every other topology visualization

Unlike threejs-network-viz, blender-3d-viz, ue5-network-viz (precise, data-driven 3D scenes) and comfyui-topology-viz (a stylized flat still image), this skill's output is a generative, non-deterministic 3D world whose geometry is not driven by the topology's actual structure — only its theme is. It is closer in spirit to concept art than to a diagram. Every result this skill produces says so explicitly and points back to the accurate diagram it was generated from (FR-009).

© automateyournetwork, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 2 other files in workspace/skills/worldlabs-topology-viz of automateyournetwork/netclaw.

  • SKILL.md
  • fantastical_prompt_builder.py
  • topology_model.py

Open the folder on GitHubat commit aa90e7d

Compare with similar skills

Worldlabs Topology Viz next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Worldlabs Topology Viz compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Worldlabs Topology Viz this skillautomateyournetwork/netclaw676—~2.9kAutomated safety check: NotesApache-2.0
Ouroboros PM InterviewQ00/ouroboros6.2k—~5.7kAutomated safety check: PassMIT
Jira Natural Language Interfacejjmartres/opencode1333 repos~1.7kAutomated safety check: PassMIT
Load Contextvishalmdi/ai-native-pm-os108—~562Automated safety check: PassNone
User Testing ValidatorIntelligent-Internet/zenith338—~1.8kAutomated safety check: PassApache-2.0
ComposerFrkAk/piyaz194—~8.6kAutomated safety check: WarnAGPL-3.0

Similar skills

  • Runs a guided product-manager interview that classifies each question automatically and produces a Product Requirements Document.

    6.2k GitHub stars~5.7k tokensUpdated 4 days ago
    Product & Project ManagementAuto-check passed
  • Lets an agent view, create, update and transition Jira issues in natural language, automatically choosing between the jira CLI and Atlassian MCP tools.

    133 GitHub starsUsed in 3 repos~1.7k tokens
    Product & Project ManagementAuto-check passed
  • Load Context

    vishalmdi/ai-native-pm-os

    Loads all PM context files and prepares Claude for a productive session.

    108 GitHub stars~562 tokensUpdated 5 mo ago
    Product & Project ManagementAuto-check passed
  • User Testing Validator

    Intelligent-Internet/zenith

    Real-surface validation coordinator for engineering validation assignments.

    338 GitHub stars~1.8k tokensUpdated 1 mo ago
    Product & Project ManagementAuto-check passed
  • Composer

    FrkAk/piyaz

    A skill your agent uses when the user types /piyaz:composer, /piyaz:composer <taskRef, or /piyaz:composer rework <taskRef|pr-url, or asks to run the next Piyaz task end-to-end, ship the backlog…

    194 GitHub stars~8.6k tokensUpdated 12 days ago
    Product & Project ManagementAuto-check: warnings
  • Cas Supervisor Checklist

    codingagentsystem/cas

    Quick startup checklist for factory supervisors. An agent skill from codingagentsystem/cas.

    176 GitHub stars~349 tokensUpdated 7 mo ago
    Product & Project ManagementAuto-check passed

More from automateyournetwork/netclaw

All 120 skills in this repo
  • EVE-NG Lab Topology Design

    automateyournetwork/netclaw

    Entry point for designing EVE-NG network labs: classifies the request, gathers missing requirements, proposes options and validates the resulting topology.

    677 GitHub stars~612 tokensUpdated today
    Auto-check passed
  • ACI Policy Change Deployment

    automateyournetwork/netclaw

    Deploys Cisco ACI policy changes only behind an approved ServiceNow Change Request, capturing pre and post-change fault baselines and rolling back automatically on a fault delta.

    677 GitHub stars~4.2k tokensUpdated today
    Auto-check passed
  • Cisco ACI Fabric Health Audit

    automateyournetwork/netclaw

    Runs a phased health audit of a Cisco ACI fabric through MCP tools: node status, links, tenant and policy review, faults and endpoint learning.

    677 GitHub stars~2.9k tokensUpdated today
    Auto-check passed
  • Anta Validation

    automateyournetwork/netclaw

    Validate Arista EOS network state against ANTA's pre-built 208-test catalogue, with structured pass/fail verdicts.

    677 GitHub stars~1.2k tokensUpdated today
    Auto-check passed
  • Arista Cvp

    automateyournetwork/netclaw

    Arista CloudVision Portal (CVP) automation via REST API — device inventory, events, connectivity monitoring, tag management (4 tools).

    677 GitHub stars~2.2k tokensUpdated today
    Auto-check: notes
  • AWS Cloud Monitoring

    automateyournetwork/netclaw

    AWS CloudWatch monitoring — metrics, alarms, log queries, VPC flow log analysis, network performance.

    677 GitHub stars~1k tokensUpdated today
    Auto-check passed

Questions about Worldlabs Topology Viz

What does Worldlabs Topology Viz do?

Turn a real network topology into a free, no-cost themed prompt preview, and optionally (after explicit confirmation) generate an explorable, AI-generated 3D 'world' via World Labs Marble — a…. Worldlabs Topology Viz is an agent skill from automateyournetwork/netclaw. Turn a real network topology into a free, no-cost themed prompt preview, and optionally (after explicit confirmation) generate an explorable, AI-generated 3D 'world' via World Labs Marble — a decorative companion visualization, never a substitute for the accurate topology diagram it is derived from.

When should I use Worldlabs Topology Viz?

Worldlabs Topology Viz fits situations like: the operator asks for a fantastical; AI-generated 3D world/scene of a network topology.

How do I install Worldlabs Topology Viz in Claude Code?

Run `npx skills add automateyournetwork/netclaw --skill worldlabs-topology-viz -a claude-code`. Or copy the skill folder (workspace/skills/worldlabs-topology-viz in automateyournetwork/netclaw) into .claude/skills/worldlabs-topology-viz in your project. Claude Code loads it when a task matches its description.

How do I install Worldlabs Topology Viz in Codex?

Run `npx skills add automateyournetwork/netclaw --skill worldlabs-topology-viz -a codex`. Or copy the skill folder (workspace/skills/worldlabs-topology-viz in automateyournetwork/netclaw) into .agents/skills/worldlabs-topology-viz in your project. Codex loads it when a task matches its description.

Can I use Worldlabs Topology Viz in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add automateyournetwork/netclaw --skill worldlabs-topology-viz -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/worldlabs-topology-viz, .gemini/skills/worldlabs-topology-viz, .github/skills/worldlabs-topology-viz and .opencode/skills/worldlabs-topology-viz in your project.

What does Worldlabs Topology Viz need to run?

Going by SKILL.md and its folder, Worldlabs Topology Viz needs Python for the scripts in its folder and credentials named WLT_API_KEY. Our summary lists: Python 3; A credential in WLT_API_KEY.

Does Worldlabs Topology Viz access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Worldlabs Topology Viz safe to install?

Our automated static check of SKILL.md found notes only (mentions a .env file), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.

What licence does Worldlabs Topology Viz use?

Worldlabs Topology Viz is published under the Apache-2.0 licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Worldlabs Topology Viz use?

About 2.9k tokens (SKILL.md is roughly 11k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Worldlabs Topology Viz?

Skills that share tags, products or a category with Worldlabs Topology Viz: Ouroboros PM Interview (Q00/ouroboros, 6.2k stars), Jira Natural Language Interface (jjmartres/opencode, 133 stars), Load Context (vishalmdi/ai-native-pm-os, 108 stars) and User Testing Validator (Intelligent-Internet/zenith, 338 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Worldlabs Topology Viz?

automateyournetwork (a GitHub user) maintains it in automateyournetwork/netclaw, which has 676 GitHub stars. The repository holds 120 skills in this directory. The repository was last updated on October 9, 2026.

Source: automateyournetwork/netclaw on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.