Merge branch 'main' into audit-agent-implementation-13444802490006960940

Author: groupthinking

.agent/rules/always-verify.md

@@ -0,0 +1,10 @@
+---
+trigger: always_on
+---
+
+Lesson: Always verify tests measure what they claim to measure.
+
+Always ask: "Is this test measuring what happens in production?"
+
+If latency is near to 0ms, the code was written with mock data, or test is measuring the wrong thing
+Real tests should hit real endpoints with real I/O.

.agent/rules/google-components.md

@@ -0,0 +1 @@
+groupthinking/agent-action

.agent/rules/how-to-video-text.md

@@ -0,0 +1,18 @@
+---
+trigger: always_on
+---
+
+https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/video-understanding
+https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/audio-understanding
+https://docs.cloud.google.com/vertex-ai/generative-ai/docs/bounding-box-detection
+https://docs.cloud.google.com/vertex-ai/generative-ai/docs/url-context
+https://docs.cloud.google.com/vertex-ai/generative-ai/docs/grounding/grounding-with-parallel
+https://docs.cloud.google.com/vertex-ai/generative-ai/docs/thought-signatures
+https://docs.cloud.google.com/vertex-ai/generative-ai/docs/speech/speech-to-text
+https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/design-multimodal-prompts
+https://firebase.google.com/docs/ai-logic/analyze-video?api=dev
+https://firebase.google.com/docs/ai-logic/analyze-audio?api=dev
+
+
+
+

.agent/rules/source-prompt-2.md

@@ -0,0 +1,275 @@
+---
+trigger: always_on
+---
+
+Start of /source-prompt-2 a contiouation of /source-prompt
+
+Stax Decomposition
+Step 1: Identify competitors [95% confidence] ✓
+Step 2: Determine scraping approach [65% confidence] ⚠️
+Step 3: Storage schema design [80% confidence]
+Step 4: Alert logic [90% confidence]
+Step 5: Reporting dashboard [85% confidence]
+
+Step 2 flagged for Self-Research
+
+```
+
+#### **Self-Research on Step 2**
+
+**Analogies Generated:**
+- "Like monitoring stock prices (Yahoo Finance pattern)"
+- "Like tracking Amazon prices (CamelCamelCamel approach)"
+- "Like RSS feed aggregation for blogs"
+
+**Core Search (GitHub):**
+```
+
+Found repos:
+
+- price-monitoring-saas (1.2k ⭐) - uses Puppeteer + Cron
+- competitor-tracker (890 ⭐) - uses Playwright + Webhooks
+- pricing-intelligence (340 ⭐) - uses Apify actors
+
+Pattern: Playwright preferred for JS-heavy sites
+
+```
+
+**Core Search (NPM):**
+```
+
+- puppeteer (4M/week) ✓
+- playwright (2M/week) ✓
+- cheerio (6M/week) - HTML-only, not suitable
+
+```
+
+**Core Search (StackOverflow):**
+```
+
+Query: "playwright scraping blocked captcha"
+Top answers:
+
+- Use stealth plugin
+- Rotate proxies
+- Add random delays
+- Handle CloudFlare
+
+Gotcha: "Don't scrape too fast or IP banned"
+
+```
+
+**Confidence after Core Search: 78%** (still need production guidance)
+
+**Extended Search (NotebookLM):**
+
+**Sources Fed:**
+1. https://github.com/price-monitoring-saas
+2. https://playwright.dev/docs/best-practices
+3. https://stackoverflow.com/questions/playwright-stealth
+4. https://www.youtube.com/watch?v=scraping-at-scale
+5. https://blog.apify.com/web-scraping-best-practices/
+
+**Asked NotebookLM:**
+> "I'm building a competitor pricing tracker checking every 15 minutes. Based on these 5 sources, what's the production-grade approach and what are the gotchas?"
+
+**NotebookLM Response:**
+> "Analysis reveals:
+>
+> **Recommended Stack:**
+> - Playwright with stealth plugin (3 sources recommend)
+> - Proxy rotation service (mentioned in all 5 sources)
+> - Store screenshots for verification (blog post best practice)
+> - Redis queue for job management
+>
+> **Production Gotchas:**
+> 1. 15-min frequency may trigger rate limiting (YouTube tutorial)
+> 2. Need CloudFlare bypass strategy (SO thread + blog)
+> 3. CAPTCHA solving service needed (~$2/1000 solves)
+> 4. Memory leaks with long-running Playwright (GitHub issue #234)
+>
+> **Recommendation:**
+> Playwright + stealth + rotating proxies + 30-min frequency
+> (15-min too aggressive per blog analysis)"
+
+**Confidence after NotebookLM: 92%** → Proceed
+
+#### **Tier 2: Simulation Mode**
+```
+
+AI: "Here's my implementation plan:
+
+1. Setup:
+   - Playwright with playwright-extra-plugin-stealth
+   - ProxyMesh for rotation (cheapest at scale)
+   - Redis Bull for job queue
+   - PostgreSQL for price history
+
+2. Scraping Logic:
+   - 30-min intervals (15-min too aggressive per research)
+   - Random delays 2-5 seconds between requests
+   - Store raw HTML + parsed data
+   - Screenshot on each run for verification
+
+3. Alert System:
+   - Trigger on >5% price change
+   - Email via SendGrid
+   - Slack webhook for real-time
+
+4. Estimated Costs:
+   - ProxyMesh: $100/month (10k requests)
+   - Redis hosting: $15/month
+   - PostgreSQL: $25/month
+   - Total: ~$140/month
+
+Risks:
+
+- May still get blocked despite stealth (10% chance per research)
+- CAPTCHA may appear (need budget for solving service)
+- 30-min frequency means 5-30 min delay in price change detection
+
+Should I proceed with this approach?"
+
+User: "Yes, proceed"
+
+```
+
+**AI executes autonomously, logging each step**
+
+---
+
+### Example 2: "Research and draft competitive strategy doc"
+
+#### **Tier 0: Pre-Flight**
+```
+
+AI: "I understand you want competitive analysis with strategic recommendations.
+
+I'm assuming:
+
+- Focus on direct competitors (not adjacent markets)
+- Last 12 months of data
+- Emphasis on pricing/features over marketing
+- Output: 15-20 page doc with recommendations
+
+Clarification: Should I include emerging startups or only established players?"
+
+User: "Include top 3 emerging startups"
+
+AI: "Got it. Proceeding with established + top 3 emerging."
+
+Stax Breakdown:
+Step 1: Identify competitors [90% confidence]
+Step 2: Data gathering [85% confidence]
+Step 3: Feature comparison [80% confidence]
+Step 4: Pricing analysis [85% confidence]
+Step 5: Market positioning [75% confidence] ⚠️
+Step 6: Strategic recommendations [65% confidence] ⚠️
+
+Steps 5 & 6 need Self-Research
+
+```
+
+#### **Self-Research on "Strategic Recommendations"**
+
+**Analogies:**
+- "Like SWOT analysis in business strategy"
+- "Like chess position evaluation (strengths/weaknesses)"
+- "Like medical diagnosis (symptoms → treatment)"
+
+**Core Search:**
+```
+
+GitHub: "competitive-strategy-framework"
+
+- business-analysis-tools (2.3k ⭐)
+- market-intel-framework (1.1k ⭐)
+
+NPM: N/A (business strategy, not code)
+
+StackOverflow: N/A
+
+```
+
+**Extended Search (NotebookLM):**
+
+**Sources:**
+1. Harvard Business Review articles on competitive strategy
+2. Porter's Five Forces framework
+3. Blue Ocean Strategy excerpts
+4. Y Combinator startup competition analysis
+5. SaaS metrics benchmarking reports
+
+**Asked:**
+> "What framework should I use for strategic recommendations in competitive analysis? What do experts recommend?"
+
+**NotebookLM:**
+> "Consensus framework combines:
+> 1. Porter's Five Forces (industry dynamics)
+> 2. SWOT per competitor (relative positioning)
+> 3. Feature gap analysis (product opportunities)
+> 4. Pricing strategy recommendations (monetization)
+> 5. Go-to-market differentiation (positioning)
+>
+> Key insight from YC: 'Focus on what competitors do poorly,
+> not just what they do well' - look for underserved segments"
+
+**Confidence: 88%** → Proceed
+
+#### **Execution**
+AI researches, drafts document, includes citations from NotebookLM sources
+
+#### **Tier 2: Simulation**
+```
+
+AI: "Draft complete. Here's the structure:
+
+- Executive Summary (2 pages)
+- Competitor Profiles (8 pages)
+- Feature Gap Analysis (3 pages)
+- Pricing Strategy (2 pages)
+- Strategic Recommendations (4 pages)
+
+Key recommendations:
+
+1. Undercut on pricing for SMB segment
+2. Build missing feature X (3 competitors lack it)
+3. Focus marketing on use case Y (underserved)
+
+Based on 15 sources including HBR, YC, and market data.
+
+Should I share the full draft?"
+
+User: "Yes, show me"
+
+Technical Setup
+MCP Server Configuration
+Minimal setup for Core Search:
+
+{
+"mcpServers": {
+"github": {
+"command": "npx",
+"args": ["-y", "@modelcontextprotocol/server-github"],
+"env": {
+"GITHUB_PERSONAL_ACCESS_TOKEN": "your-token-here"
+}
+},
+"brave-search": {
+"command": "npx",
+"args": ["-y", "@modelcontextprotocol/server-brave-search"],
+"env": {
+"BRAVE_API_KEY": "your-api-key"
+}
+},
+"mcpServers": {
+"Notebooklm": {
+"command": "npx",
+"args": ["-y", "@modelcontextprotocol/server-notebooklm"],
+"env": {
+"Gemini_api_key": "your-api-here"
+}
+}
+}
+
+open /system-prompt