v1.1.0 to v1.2.0
27 added, 33 removed. Audit A to A.
---
name: humanization-stealth-browsing
- description: >
- Browses web with human-like patterns to reduce bot detection. Use for scrapi
- ngfragile sites or user says stealth browse, human-like browsing. Use when t
- heuser needs this capability. Triggers: stealth browse, human-like browsing,
- avoid bot detection.
- version: 1.1.0
+ description: "Apply respectful, rate-limited browsing practices for public websites while honoring site rules and access controls. Use for: responsible web research, rate limiting, CAPTCHA stop, robots.txt."
+ version: 1.2.0
author: Stijnman
license: MIT
+ compatibility: Grok agent; optional browser and shell access
metadata:
grok:
- tags: [stealth browse, human-like browsing, avoid bot detection]
- related_skills: [web-scraper, sandbox-internet-handler, humanization-stealth-browsing]
- compatibility: Grok agent; optional MCP and shell access
+ tags: [responsible browsing, rate limiting, public web research, robots.txt, CAPTCHA stop]
+ related_skills: [web-scraper, sandbox-internet-handler]
---
- # Humanization Stealth Browsing
+ # Responsible Web Browsing
## When to Use
- - User says **stealth browse** or task matches this capability
- - User says **human-like browsing** or task matches this capability
- - User says **avoid bot detection** or task matches this capability
+ Use this skill when collecting information from public websites requires careful pacing, stable navigation, or respect for fragile endpoints. It supports responsible research, not bypassing protections.
+ > Do not use this skill to evade bot detection, bypass access controls, defeat CAPTCHAs, circumvent paywalls, or access content that the user is not authorized to view.
+
## Workflow
- 1. Set realistic delays and headers.
- 2. Navigate incrementally; avoid burst requests.
- 3. Rotate user-agent only when site blocks.
- 4. Respect robots.txt; stop on CAPTCHA and ask user.
+ 1. Confirm that the target is public and that the requested collection is legitimate.
+ 2. Prefer official APIs, published feeds, or a site’s documented export options.
+ 3. Review and respect applicable terms, `robots.txt`, rate limits, and access restrictions.
+ 4. Make requests incrementally, cache results where appropriate, and avoid unnecessary retries.
+ 5. If a login wall, CAPTCHA, paywall, or block appears, stop and explain the limitation to the user.
- ## Integrations
+ ## Operating Principles
- - `web-scraper`
- - `sandbox-internet-handler`
- - `humanization-stealth-browsing`
+ | Principle | Practice |
+ |---|---|
+ | Minimize load | Use measured request rates and avoid broad, repetitive crawling. |
+ | Preserve provenance | Record source URLs and retrieval dates for material findings. |
+ | Respect controls | Do not rotate identities, automate CAPTCHA solving, or bypass authentication. |
+ | Fail safely | Stop rather than escalating when a site signals restricted access. |
## Error Handling
- | Failure | Response |
- |---------|----------|
- | CAPTCHA encountered | Stop; ask user to solve manually. |
-
- ## Gotchas
-
- - Never bypass paywalls or auth without permission.
-
- ## Example
-
- **Input:** User request matching triggers above.
- **Output:** Structured result per workflow with integrations invoked as needed.
+ | Situation | Response |
+ |---|---|
+ | Rate limit or temporary block | Pause activity and offer a slower or official-data-source alternative. |
+ | CAPTCHA or login wall | Stop the automated flow; ask the user to provide authorized access if appropriate. |
+ | `robots.txt` or terms restrict collection | Do not scrape the restricted path; propose permitted alternatives. |
+ | Source data is incomplete | State the limitation and avoid filling gaps with assumptions. |