snaprender

A web screenshot and page-content extraction tool for coding agents. It can capture pages on simulated phones, tablets, or computers and return text, Markdown, HTML, or structured data.

In plain words
What is it for?
Use it to take screenshots, compare desktop and mobile layouts, extract articles, or collect links from a URL.
Why use it?
It helps an agent inspect how a webpage looks and what it contains when text access alone is not enough.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/user0856/snaprender-integrations/openclaw
Any agent
npx skills add User0856/snaprender-integrations --skill openclaw
Clone the repo
git clone --depth 1 https://github.com/User0856/snaprender-integrations

Made for: Claude Code, Codex.

Per session 82 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,356 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 2 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00082 $0.03356
Opus 5 $0.00041 $0.01678
Sonnet 5 $0.00016 $0.00671
Haiku 4.5 $0.00008 $0.00336

Measured yesterday against content hash d2de6a6a19ac, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

snaprender scanned grade B with 2 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 yesterday.

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.

Sends data to an external URLmediumData exfiltration

A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.

jq -n --arg url 'https://stripe.com' '{url: $url, response_type: "json", format: "jpeg", quality: 60, block_ads: true, block_cookie_banners: true}' | curl -s -X POST "https://app.snap-render.com/v1/screenshot" -H "X-API-

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

description: "Give your agent eyes on the web — screenshot any URL as an image file, or extract page content as markdown/text/HTML. Supports device emulation (iPhone, iPad, Pixel, MacBook), dark mode, full-page scroll, a
openclaw/SKILL.md · 258 lines

How it starts

The opening of the file, as written. The whole thing — 258 lines — stays where its author put it; the contents beside it link to each section on GitHub.

SnapRender — Give Your Agent Eyes

Your agent can read the web but can't see it. One command and it captures pixel-perfect screenshots or extracts page content as clean markdown, plain text, or structured data.

"Screenshot stripe.com on iPhone", "Extract the article from this blog post", "Compare desktop vs mobile", "Get all links from this page" — just ask.

Free tier: 200 requests/month, no credit card. Get a key ->


IMPORTANT: Use the exec tool with curl. NEVER use the browser tool for screenshots or extraction.


Screenshot: How to Capture

Run this command via the exec tool. Replace TARGET_URL with the target URL:

jq -n --arg url 'TARGET_URL' \
  '{url: $url, response_type: "json", format: "jpeg", quality: 60, block_ads: true, block_cookie_banners: true}' \
| curl -s -X POST "https://app.snap-render.com/v1/screenshot" \
  -H "X-API-Key: $SNAPRENDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d @- \
| tee /tmp/snap_response.json \
| jq -r '.image' | sed 's|data:image/[^;]*;base64,||' | base64 -d > /tmp/screenshot.jpg \
&& jq '{url, format, size, cache, responseTime, remainingCredits}' /tmp/snap_response.json

This saves the screenshot to /tmp/screenshot.jpg and prints metadata.

Screenshot: Rules

  1. Use exec tool only — NEVER the browser tool
  2. $SNAPRENDER_API_KEY is already set — use it literally in the command, do NOT replace it
  3. Always build JSON with jq --arg — never interpolate user input directly into shell strings or JSON. Pass values via jq -n --arg to prevent injection
  4. Always use format=jpeg&quality=60 — keeps response small enough for the agent context
  5. Always pipe to save the image to a file — the base64 response is too large to display inline
  6. Report metadata to the user — file size, response time, cache status, remaining credits

Render HTML or Markdown

Use POST with a JSON body to render raw HTML or Markdown content (no URL needed). Always use jq --arg to safely pass content:

Read the full file on GitHub · 258 lines

Files

What ships with it

1 file 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.

Changes

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.

  1. yesterday First seen · 258 lines · 82 tokens per session scan B d2de6a6a19ac

Subscribe to this mod's changes

snaprender is a skill published in the GitHub repository User0856/snaprender-integrations (2 stars, last pushed 2mo ago), licensed MIT. It adds 82 tokens to every session and 3,356 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B with 2 findings (sends data to an external url, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.