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 AllenS0104/skill-browser --skill external-site-profile-learninggit clone --depth 1 https://github.com/AllenS0104/skill-browserWrote 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/allens0104/skill-browser/external-site-profile-learning)<a href="https://agentmods.dev/skills/allens0104/skill-browser/external-site-profile-learning"><img src="https://agentmods.dev/badge/skills/allens0104/skill-browser/external-site-profile-learning/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/allens0104/skill-browser/external-site-profile-learning"><img src="https://agentmods.dev/badge/skills/allens0104/skill-browser/external-site-profile-learning.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00054 | $0.01981 |
| Opus 5 | $0.00027 | $0.00991 |
| Sonnet 5 | $0.00011 | $0.00396 |
| Haiku 4.5 | $0.00005 | $0.00198 |
Grade A, and why
external-site-profile-learning 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 9d 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 — 242 lines — stays where its author put it; the contents beside it link to each section on GitHub.
External site profile learning skill
Use this skill when working on the target Playwright browser project (for example, the 99idea demo project).
This project is a Playwright-based browser demo that supports config-driven external site profiles through:
scenarios\site-profiles.jsonsrc\site-profiles.jssrc\planner.jssrc\llm-planner.jssrc\run-demo.js
What this skill is for
Use this skill when the task involves any of the following:
- adding a new external site profile
- debugging why an external site passes heuristics but fails with Gemini
- turning a random website experiment into a stable config-driven profile
- identifying whether a site is suitable for the default regression matrix
- distinguishing selector problems from login, anti-bot, or hidden-UI problems
Core rules
- Prefer URL or title verification for external sites whenever possible.
- Prefer stable, unique selectors over generic selectors like
buttonor[name="q"]. - If a site profile exists, the final plan should use the profile's exact
searchInputSelectorandsubmitSelector. - Keep GitHub out of the default regression matrix. It works, but repeated probing can trigger rate limits.
- Classify the site's execution tier before writing a browser profile:
- If opencli already supports the site, prefer an opencli-based workflow over a selector-based profile
- If the site has a public API, prefer fetch over browser automation
- Only write a browser profile when no higher-tier execution path exists
- Treat these as different failure classes:
- selector ambiguity
- hidden or disabled controls
- opener-required search UI
- SPA shell (no usable DOM without full JS rendering)
- login gates
- anti-bot / rate-limit behavior
- heavy pages that stall on
load
Dual-engine awareness (inspired by opencli)
When stabilizing a site, consider whether the profile should be:
- Data-only (YAML-like): selector + verification config in JSON — for traditional DOM-based sites
- Runtime-aware: includes execution path recommendation (opencli / fetch / browser) — for modern platforms
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.
- 9d ago First seen · 242 lines · 54 tokens per session scan A 3caa4ec2436d
external-site-profile-learning is a skill published in the GitHub repository AllenS0104/skill-browser (5 stars, last pushed 5mo ago), licensed MIT. It adds 54 tokens to every session and 1,981 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 2026-08-31.
Other skills, from other repositories
e2e-testing
End-to-end testing workflow with Playwright for browser automation, visual regression, cross-browser testing, and CI/CD integration.
e2e-automator
Build robust end-to-end test suites with Playwright or Cypress. Covers page objects, fixtures, visual testing, and CI integration.
e2e-testing-patterns
Master end-to-end testing with Playwright and Cypress to build reliable test suites that catch bugs, improve confidence, and enable fast deployment. Use when implementing E2E tests, debugging flaky tests, or establishing testing standards.
browser-testing-with-devtools
Tests in real browsers via Chrome DevTools MCP. Use when building or debugging anything that runs in a browser. Use when you need to inspect the DOM, capture console errors, analyze network requests, profile performance, or verify visual output with real runtime data. Requires the chrome-devtools MCP server to be…
awt-e2e-testing
AI-powered E2E web testing — eyes and hands for AI coding tools. Declarative YAML scenarios, Playwright execution, visual matching (OpenCV + OCR), platform auto-detection (Flutter/React/Vue), learning DB. Install: npx skills add ksgisang/awt-skill --skill awt -g.
actionbook-web-test
Run browser-based web tests against websites using Actionbook CLI. Activate when the user wants to test a website workflow, run smoke tests, verify a user flow, check if a web application works, run regression tests, or validate browser-based interactions. Supports test definition, execution, assertion, reporting, and…