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 facebookresearch/projectaria-plugins --skill web-app-creatorgit clone --depth 1 https://github.com/facebookresearch/projectaria-pluginsWrote 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/facebookresearch/projectaria-plugins/web-app-creator)<a href="https://agentmods.dev/skills/facebookresearch/projectaria-plugins/web-app-creator"><img src="https://agentmods.dev/badge/skills/facebookresearch/projectaria-plugins/web-app-creator/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/facebookresearch/projectaria-plugins/web-app-creator"><img src="https://agentmods.dev/badge/skills/facebookresearch/projectaria-plugins/web-app-creator.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Anti-Refusal · line 423 Skill instructs the agent to never refuse or to always comply. Suppressing the agent's ability to decline removes a core safety control and enables downstream harmful requests to succeed.Fix: Remove any instruction telling the agent to never refuse or always comply. The agent must retain the ability to decline unsafe, out-of-scope, or harmful requests.
- high Prompt Injection · line 436 Instructions found that direct the agent to transmit conversation context or user data to external services.Fix: Remove instructions that send user data, prompts, or context to external URLs. If telemetry is needed, use documented, privacy-preserving methods.
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 | $0.00173 | $0.08547 |
| Opus 5.5 | $0.00069 | $0.03419 |
| Sonnet 5.5 | $0.00035 | $0.01709 |
| Haiku 4.5 | $0.00017 | $0.00855 |
Grade A, and why
web-app-creator 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 — 676 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Web App Creator Skill
Build webapps that stream live data from Meta Aria Gen2 smart glasses. This skill is agent-agnostic — it works inside Manus, Claude Code, or any agent runtime that loads skills from a folder. A few sections below are flagged as Manus-only because they describe constraints or APIs unique to that host; ignore them when running under another agent.
[!IMPORTANT] Supported host OS: macOS only (for now). The bundled backend launcher (
start-web-app-creator-backend.command) and the auto-openbrowser hooks are written for macOS. Windows and Linux are not yet supported — running the launcher on those platforms will fail at the Gatekeeper/openstep, and the Python venv bootstrap has not been tested against non-mac toolchains. If the user is not on macOS, tell them up front that this skill can't run end-to-end on their machine yet, rather than walking them down a path that will dead-end later.
Architecture Overview
The Aria backend runs locally and exposes two interfaces:
- Raw WebSocket —
ws://localhost:17300— push-based JSON text + binary image frames, server → client - nebulaDataUpdate event bridge — the Web App Creator frontend re-dispatches WebSocket messages as DOM
CustomEvents so generated HTML components can consume data without managing their own WebSocket connection
Which pattern to use:
| Use Case | Pattern |
|---|---|
| Standalone webapp / new project | Raw WebSocket (ws://localhost:17300) |
| Component generated inside Web App Creator iframe | nebulaDataUpdate event listener |
| Full voice pipeline (STT + LLM + TTS) | Requires full-stack project (web-db-user) |
Starting the Backend
The Web App Creator backend is what serves ws://localhost:17300 and what every template in this skill connects to. Nothing in the skill works until this is running.
The only supported way: double-click the launcher
Tell the user to:
- Open Finder, navigate to the skill folder (
skills/web-app-creator/), then into theweb-app-creator-backend/subfolder - Double-click
start-web-app-creator-backend.command - A Terminal window opens, the colored launcher banner appears, and it prints
WebSocket listening on ws://localhost:17300once the device is streaming - Leave that Terminal window open while you use the skill. To stop, press
Ctrl+Cin that window (or just close it)
What ships with it
31 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- references/sensor_schemas.md 8.3 KB
- references/visual_style_guide.md 14 KB
- references/whisper_hallucinations_notes.txt 1.0 KB
- templates/device-pose-and-hands/aria-icon-dark.png 26 KB
- templates/device-pose-and-hands/DevicePoseAndHands.html 71 KB
- templates/device-pose-and-hands/README.md 1.4 KB
- templates/resources/aria-glasses.obj 2863 KB
- templates/resources/aria-icon-dark.png 26 KB
- templates/resources/aria-specs.png 6.4 KB
- templates/resources/glasses_wireframe.json 192 KB
- templates/resources/glasses.json 55 KB
- templates/stream-panels/README.md 2.1 KB
- templates/stream-panels/StreamPanel.html 53 KB
- templates/voice-pipeline/.env.example 3.5 KB
- templates/voice-pipeline/client/useAriaAudio.ts 4.3 KB runs code
- templates/voice-pipeline/client/useVAD.ts 12 KB runs code
- templates/voice-pipeline/examples/express-server.ts 2.8 KB runs code
- templates/voice-pipeline/examples/manus-trpc-router.ts 11 KB runs code
- templates/voice-pipeline/README.md 9.6 KB
- templates/voice-pipeline/server/hallucinationFilter.ts 5.4 KB runs code
- templates/voice-pipeline/server/index.ts 1.1 KB runs code
- templates/voice-pipeline/server/llmChat.ts 7.1 KB runs code
- templates/voice-pipeline/server/transcribeAudio.ts 8.2 KB runs code
- templates/voice-pipeline/server/types.ts 2.6 KB runs code
- web-app-creator-backend/backend/__init__.py 743 B runs code
- web-app-creator-backend/backend/websocket_bridge.py 110 KB runs code
- web-app-creator-backend/profiles/web_app_creator_full.json 1.3 KB
- web-app-creator-backend/profiles/web_app_creator_light.json 814 B
- web-app-creator-backend/profiles/web_app_creator_standard.json 858 B
- web-app-creator-backend/requirements.txt 458 B
- web-app-creator-backend/start-web-app-creator-backend.command 49 KB
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 · 676 lines · 173 tokens per session scan A 809d8f24d274
web-app-creator is a skill published in the GitHub repository facebookresearch/projectaria-plugins (14 stars, last pushed 6d ago), licensed Apache-2.0. It adds 173 tokens to every session and 8,547 once invoked, about $0.0007 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 2026-08-31.
Other skills, from other repositories
gke-compute-classes
Configures, optimizes, and troubleshoots GKE ComputeClasses. Use when configuring Spot VMs with on-demand fallback, targeting specific accelerators (GPUs/TPUs) or machine families, restricting ComputeClass access, or debugging pending pods related to node pool auto-creation. Do not use for cluster-level Node Auto…
jetson-diagnostic
Read-only Jetson health snapshot for identity, memory, GPU, thermal, power, storage, services, and top processes.
offensive-z-wave
Z-Wave attack methodology — sniffing with Z-Force / EZ-Wave / RTL-SDR + ZniffMobile, S0 (legacy) network-key derivation flaw and key reuse, S2 (modern) ECDH commissioning analysis, replay/injection on unauthenticated nodes, default-key brute-force on test deployments, and home-automation hub pivots. Use when targeting…
doca-socket-relay
Use this skill when the operator is driving the DOCA Socket Relay to bridge a socket-oriented host application onto a BlueField DPU peer without rewriting it — picking the deployment shape (in-process, sidecar, or BlueField service container), configuring the host-side socket and the DPU-side forwarding endpoint…
hsb-flash
Flash the FPGA on an HSB board connected to an NVIDIA devkit. Supports HSB Lattice boards (FPGA versions 2407, 2412, 2507, 2510) and Leopard Imaging VB1940 "all-in-one" cameras (FPGA versions 2507, 2510). Uses release-specific YAML manifests and board-type-specific program commands. Lattice and VB1940 commands must…
jetson-validate-image
Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.