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/softaworks/agent-toolkit/web-to-markdownnpx skills add softaworks/agent-toolkit --skill web-to-markdowngit clone --depth 1 https://github.com/softaworks/agent-toolkitWrote 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/softaworks/agent-toolkit/web-to-markdown)<a href="https://agentmods.dev/skills/softaworks/agent-toolkit/web-to-markdown"><img src="https://agentmods.dev/badge/skills/softaworks/agent-toolkit/web-to-markdown.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 | $0.00064 | $0.00944 |
| Opus 5 | $0.00032 | $0.00472 |
| Sonnet 5 | $0.00013 | $0.00189 |
| Haiku 4.5 | $0.00006 | $0.00094 |
Grade A, and why
web-to-markdown 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 4d 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 — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.
web-to-markdown
Convert web pages to clean Markdown by driving a locally installed browser (via web2md).
Hard trigger gate (must enforce)
This skill MUST NOT be used unless the user explicitly wrote exactly a phrase like:
use the skill web-to-markdown ...use a skill web-to-markdown ...
If the user did not explicitly request this skill by name, stop and ask them to re-issue the request including: use the skill web-to-markdown.
What this skill does
- Handles JS-rendered pages (Puppeteer → user Chrome).
- Works best with Chromium-family browsers (Chrome/Chromium/Brave/Edge) via
puppeteer-core. - Extracts main content (Readability).
- Converts to Markdown (Turndown) with cleaned links and optional YAML frontmatter.
Non-goals
- Do not use Playwright or other browser automation stacks; the mechanism is
web2md.
Inputs you should collect (ask only if missing)
url(or a list of URLs)- Output preference:
- Print to stdout (
--print), OR - Save to a file (
--out ./file.md), OR - Save to a directory (
--out ./some-dir/to auto-name by page title)
- Print to stdout (
- Optional rendering controls for tricky pages:
--chrome-path <path>(if Chrome auto-detection fails)--interactive(show Chrome and pause so the user can complete human checks/login, then press Enter)--wait-until load|domcontentloaded|networkidle0|networkidle2--wait-for '<css selector>'--wait-ms <milliseconds>--headful(debug)--no-sandbox(sometimes required in containers/CI)--user-data-dir <dir>(login/session; use a dedicated profile directory)
Workflow
- Confirm the user explicitly invoked the skill (
use the skill web-to-markdown). - Validate URL(s) start with
http://orhttps://. - Ensure
web2mdis installed:- Run:
command -v web2md - If missing, instruct the user to install it (assume the project exists at
~/workspace/softaworks/projects/web2md):cd ~/workspace/softaworks/projects/web2md && npm install && npm run build && npm link- Or:
cd ~/workspace/softaworks/projects/web2md && npm install && npm run build && npm install -g .
- Run:
- Convert:
- Single URL → file:
web2md '<url>' --out ./page.md
- Single URL → auto-named file in directory:
mkdir -p ./out && web2md '<url>' --out ./out/
- Human verification / login walls (interactive):
mkdir -p ./out && web2md '<url>' --interactive --user-data-dir ./tmp/web2md-profile --out ./out/- Then: complete the check in the browser window and press Enter in the terminal to continue.
- Print to stdout:
web2md '<url>' --print
- Multiple URLs (batch):
- Create output dir (e.g.
./out/) then run oneweb2mdcommand per URL using--out ./out/
- Create output dir (e.g.
- Single URL → file:
- Validate output:
- If writing files, verify they exist and are non-empty (e.g.
ls -la <path>andwc -c <path>).
- If writing files, verify they exist and are non-empty (e.g.
- Return:
- The saved file path(s), or the Markdown (stdout mode).
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.
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.
- 4d ago First seen · 78 lines · 64 tokens per session scan A 9c61df95ffd1
web-to-markdown is a skill published in the GitHub repository softaworks/agent-toolkit (2,422 stars, last pushed 6mo ago), licensed MIT. It adds 64 tokens to every session and 944 once invoked, about $0.0003 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 2026-08-30.
Other skills, from other repositories
webapp-testing
Start/reuse a local app, wait for readiness, inspect rendered state/console/network, act from observed selectors, and verify with evidence.
webbrowser
Use the native webbrowser module for web search, browsing, authenticated website interaction, forms, file uploads and downloads, screenshots, and user handoff.
control-chrome
Control the user's Chrome browser for tasks that depend on existing Chrome state: tabs, logged-in sessions, or extensions. Prefer purpose-built connectors, APIs, or CLIs when available.
research-agent
Use for current web/X research, library/API verification, competitive analysis, source-backed technical decisions, or updating docs with recent facts.
agent-browser
Use agent-browser CLI for website automation: navigate, click/fill, scrape, screenshot, test web apps, debug console/network.
youtube-watcher
Fetch and read transcripts from YouTube videos. Use when you need to summarize a video, answer questions about its content, or extract information from it.