aboylabs/ai-agent-skills / skills/devops/vps-cleanup
vps-cleanup
VerifiedSystematic VPS cleanup — analyze files, categorize by importance, safely delete temporary/old data to free disk space
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 14.2 KB |
| references/hermes-ecosystem-audit.md | markdown | 7.4 KB |
| references/scrapling-vs-tinyfish-comparison.md | markdown | 7.1 KB |
references/scrapling-vs-tinyfish-comparison.md
View on GitHub# Scrapling vs TinyFish — VPS Resource Comparison
**Session:** 2026-05-12
**Context:** User asked to compare Scrapling (Python web scraping framework) vs TinyFish (cloud API service) for VPS usage
## Quick Verdict
**For VPS with 2GB RAM, 40GB disk:**
- ✅ **TinyFish** — 0 MB disk, 0 MB RAM (cloud API)
- ❌ **Scrapling** — 650-900 MB disk, 200-400 MB RAM (local execution)
**User response:** "gede banget makan emori dan ram" → **INSTANT REJECTION**
## Detailed Comparison
### Architecture
| Aspect | Scrapling | TinyFish |
|--------|-----------|----------|
| **Type** | Python library/framework | Cloud API service |
| **Install** | `pip install scrapling` | Just API key |
| **Disk** | 7.0 MB (238 files, 110 Python) | 0 MB (API only) |
| **Dependencies** | Playwright/browser (300-400 MB) | None |
| **Total Disk** | **650-900 MB** | **0 MB** |
| **RAM (idle)** | 50 MB | 0 MB |
| **RAM (active)** | **200-400 MB** per session | **0 MB** |
| **Execution** | Local (your VPS) | Cloud (TinyFish servers) |
### Features
**Scrapling:**
- 4 fetchers (Fetcher, AsyncFetcher, StealthyFetcher, DynamicFetcher)
- Adaptive parsing (learns from website changes)
- Auto-save selectors (survives design changes)
- Cloudflare Turnstile bypass (built-in)
- Full spider framework (concurrent crawls)
- Proxy rotation (automatic)
- Pause/resume crawls
- Real-time stats & streaming
- CLI tool
- MCP server support
- Browser automation (Playwright-based)
**TinyFish:**
- Search API (live web search)
- Fetch API (clean text extraction)
- Browser API (Cloudflare bypass)
- Agent API (multi-step workflows)
- Batch fetch (max 10 URLs)
- Geo-targeted search
- Free tier (search + fetch)
**Winner:** Scrapling (way more features) — but irrelevant if can't run on VPS
### Resource Impact on VPS
**VPS Specs:** 2 vCPUs, 2GB RAM, 40GB disk
**Scrapling:**
```
Disk: 650-900 MB (16-22% of 40GB)
RAM: 200-400 MB per session (10-20% of 2GB)
Risk: OOM with concurrent sessions
Impact: 🔴 HIGH — user rejected immediately
```
**TinyFish:**
```
Disk: 0 MB (0% of 40GB)
RAM: 0 MB (0% of 2GB)
Risk: None (cloud execution)
Impact: 🟢 ZERO — perfect for VPS
```
### Cost
**Scrapling:**
- Library: FREE (open source)
- Proxies: You pay (if needed)
- Infrastructure: Your VPS/local machine
- Maintenance: You handle updates
**TinyFish:**
- Search: FREE ✅
- Fetch: FREE ✅
- Browser API: Paid (credits)
- Agent API: Paid (credits)
- Infrastructure: TinyFish handles
- Maintenance: TinyFish handles
**Winner:** TinyFish (free tier covers most use cases)
### Use Cases
**Scrapling best for:**
- Complex crawling projects (1000s of pages)
- Custom parsing logic
- Adaptive selectors (website changes)
- Full browser control
- Local execution (privacy)
- Powerful machine (8GB+ RAM)
**TinyFish best for:**
- Simple scraping (1-100 pages)
- Search + fetch workflow
- Resource-constrained VPS (2GB RAM)
- Quick prototyping
- API-first architecture
- Zero maintenance
## User Decision Pattern
**Question:** "Scrapling bagus gak?"
**Analysis provided:**
1. Features comparison (Scrapling wins)
2. Resource comparison (TinyFish wins)
3. Architecture comparison (different approaches)
4. Cost comparison (both have free options)
5. **Resource impact** (CRITICAL for VPS)
**User response:** "gede banget makan emori dan ram"
**Translation:** "Too big, eats memory and RAM"
**Meaning:**
- Resource cost is deal-breaker
- No discussion needed
- No workarounds wanted
- Instant rejection
**Correct response:**
- ✅ Acknowledge rejection immediately
- ✅ Recommend TinyFish (already integrated)
- ✅ Explain why TinyFish is better for VPS
- ❌ Don't offer Scrapling workarounds
- ❌ Don't suggest "minimal install"
- ❌ Don't explain Scrapling benefits
## Key Learnings
### 1. Resource Cost is Non-Negotiable
When user has VPS constraints (2GB RAM, 40GB disk):
- Resource cost > features
- 0 MB > powerful features
- Cloud API > local framework
### 2. User's Rejection Pattern
**Phrase:** "gede banget makan emori dan ram"
**Signals:**
- ZERO tolerance for heavy tools
- Resource cost is deal-breaker
- No negotiation, no workarounds
- Move on immediately
**Response:**
- Stop immediately
- Don't defend the tool
- Don't offer alternatives (minimal install, Docker, etc.)
- Recommend lightweight alternative
- Move on
### 3. Evaluation Order
When recommending tools for VPS:
**WRONG order:**
1. Features (what it can do)
2. Cost (how much it costs)
3. Resource (how much it uses) ← Too late!
**RIGHT order:**
1. **Resource** (can it run on VPS?) ← Check FIRST
2. Features (what it can do)
3. Cost (how much it costs)
**If resource check fails → STOP, find alternative**
### 4. Thresholds for VPS (2GB RAM, 40GB disk)
| Resource | Threshold | Action |
|----------|-----------|--------|
| Disk < 50 MB | ✅ Safe | Recommend without asking |
| Disk 50-500 MB | 🟡 Moderate | Ask first, explain cost |
| Disk > 500 MB | 🔴 Heavy | **Find alternative or skip** |
| RAM < 50 MB | ✅ Safe | Recommend without asking |
| RAM 50-100 MB | 🟡 Moderate | Ask first, explain cost |
| RAM > 100 MB | 🔴 Heavy | **Find alternative or skip** |
**Scrapling:** 650-900 MB disk + 200-400 MB RAM = **🔴 INSTANT REJECTION**
## Alternative Strategies
When tool is too heavy for VPS:
### 1. Cloud API (Best)
- Browserbase (cloud browser)
- TinyFish (cloud scraping)
- ScrapingBee (cloud scraping)
- **Cost:** 0 MB disk, 0 MB RAM
- **Trade-off:** API credits
### 2. Lightweight Library
- requests + BeautifulSoup (vs Scrapling)
- curl + jq (vs complex tools)
- **Cost:** < 50 MB disk, < 50 MB RAM
- **Trade-off:** Less features
### 3. Docker (Still Heavy)
- Isolated environment
- **Cost:** Still 650-900 MB disk + 200-400 MB RAM
- **Trade-off:** Isolation, but same resource cost
- **Verdict:** Still rejected by user
### 4. Skip Feature
- If not critical, skip
- **Cost:** 0 MB
- **Trade-off:** No feature
## Common Heavy Tools
| Tool | Disk | RAM (active) | Alternative |
|------|------|--------------|-------------|
| **Scrapling** | 650-900 MB | 200-400 MB | TinyFish (0 MB) |
| **Chrome/Chromium** | 300-400 MB | 150-300 MB | Browserbase (0 MB) |
| **Playwright** | 400-500 MB | 200-300 MB | Browserbase (0 MB) |
| **Selenium** | 350-450 MB | 150-250 MB | Browserbase (0 MB) |
| **Puppeteer** | 300-400 MB | 150-250 MB | Browserbase (0 MB) |
**Pattern:** Browser automation tools = heavy (300-900 MB)
**Solution:** Use cloud browser services (Browserbase, Browser Use Cloud)
## Recommendation Template
When user asks about heavy tool:
```markdown
## Tool Analysis: [TOOL_NAME]
### Resource Cost:
- Disk: [SIZE] MB
- RAM: [SIZE] MB (active)
- Total: [DISK + RAM] MB
### VPS Impact:
- Disk: [X]% of 40GB
- RAM: [X]% of 2GB
- Verdict: 🔴 TOO HEAVY / 🟡 MODERATE / 🟢 SAFE
### Alternative:
- [LIGHTWEIGHT_ALTERNATIVE]
- Disk: [SIZE] MB
- RAM: [SIZE] MB
- Trade-off: [FEATURES_LOST]
### Recommendation:
✅ Use [ALTERNATIVE] (0 MB disk, 0 MB RAM)
❌ Skip [TOOL_NAME] (too heavy for VPS)
```
## References
- Scrapling: https://github.com/D4Vinci/Scrapling
- TinyFish: https://tinyfish.ai
- Browserbase: https://browserbase.com
- Session: 2026-05-12 (user: ryzen, VPS: 2GB RAM, 40GB disk)
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |