$npx -y skills add adityahimaone/hermes-agent-rtk-caveman --skill financial_analysis_automationCreate automated financial analysis skills with scheduled execution and result storage. Use when you need to build custom stock/scanning analysis that runs automatically and saves results to Obsidian for daily review.
| 1 | # Financial Analysis Automation Skill |
| 2 | |
| 3 | Use this skill to create automated financial analysis workflows that: |
| 4 | 1. Scan multiple sources for investment opportunities |
| 5 | 2. Run on a schedule (via cron jobs) |
| 6 | 3. Save formatted results to Obsidian vault |
| 7 | 4. Can be monitored and tracked |
| 8 | |
| 9 | ## When to Use |
| 10 | - When you want to create custom stock/scanning analysis |
| 11 | - When you need analysis to run automatically on a schedule |
| 12 | - When you want results saved to your Obsidian vault for review |
| 13 | - When you need to track multiple financial data sources |
| 14 | - When you want to combine web crawling with sentiment analysis |
| 15 | |
| 16 | ## Core Components |
| 17 | |
| 18 | ### 1. Analysis Skill Creation |
| 19 | Create a custom skill that includes: |
| 20 | - **Data Sources**: Define where to get information (websites, APIs, social media) |
| 21 | - **Analysis Logic**: Define what to look for (volume spikes, sentiment, patterns) |
| 22 | - **Output Format**: Define how results should be presented |
| 23 | - **Error Handling**: Define what to do when sources fail |
| 24 | |
| 25 | ### 2. Automation Setup |
| 26 | Set up scheduled execution using: |
| 27 | - **Cron Jobs**: Schedule the skill to run at specific times |
| 28 | - **Skill Chaining**: Combine multiple skills (e.g., analysis + note saving) |
| 29 | - **Time Filtering**: Ensure only recent data is analyzed |
| 30 | - **Source Validation**: Validate data quality and freshness |
| 31 | |
| 32 | ### 3. Result Storage |
| 33 | Configure automatic saving to: |
| 34 | - **Obsidian Vault**: Save results as formatted notes |
| 35 | - **Proper Frontmatter**: Include metadata for filtering and linking |
| 36 | - **Consistent Naming**: Use date-based filenames for easy sorting |
| 37 | - **Tagging System**: Add relevant tags for discovery |
| 38 | |
| 39 | ## Workflow Steps |
| 40 | |
| 41 | ### Step 1: Define Analysis Requirements |
| 42 | - What financial instruments to analyze (stocks, crypto, etc.) |
| 43 | - What data sources to use (financial sites, social media, news) |
| 44 | - What specific metrics or patterns to look for |
| 45 | - What time sensitivity is required (real-time, daily, weekly) |
| 46 | |
| 47 | ### Step 2: Create the Analysis Skill |
| 48 | Create a custom Hermes skill that: |
| 49 | - Uses appropriate data extraction methods (firecrawl, web_search, etc.) |
| 50 | - Applies your analysis logic and filters |
| 51 | - Formats output in a consistent, readable way |
| 52 | - Includes risk disclaimers and limitations |
| 53 | |
| 54 | ### Step 3: Set Up Automation |
| 55 | Configure a cron job that: |
| 56 | - Runs your analysis skill on the desired schedule |
| 57 | - Chains with appropriate output skills (obi-notes, etc.) |
| 58 | - Includes proper error handling and logging |
| 59 | - Runs during appropriate market hours |
| 60 | |
| 61 | ### Step 4: Configure Output Storage |
| 62 | Set up the result saving to: |
| 63 | - Create appropriate folder structure in Obsidian |
| 64 | - Generate properly formatted markdown with frontmatter |
| 65 | - Include metadata about the analysis (time, sources, etc.) |
| 66 | - Add relevant tags for easy discovery and filtering |
| 67 | |
| 68 | ### Step 5: Monitor and Maintain |
| 69 | - Check that the cron job runs successfully |
| 70 | - Verify output is correctly formatted and saved |
| 71 | - Update analysis logic as market conditions change |
| 72 | - Refresh API keys and credentials as needed |
| 73 | |
| 74 | ## Example: Market Alpha Scout |
| 75 | The market_alpha_scout skill demonstrates this workflow: |
| 76 | - **Analysis Skill**: market_alpha_scout (research category) |
| 77 | - **Data Sources**: |
| 78 | - Stockbro.id (firecrawl crawl with Google Workspace auth) |
| 79 | - Financial news sites (Google search: detik, kontan, bisnis) |
| 80 | - Social sentiment (Google search: X/Twitter content like @happenfed) |
| 81 | - **Schedule**: Weekdays at 08:15 WIB (before market open) |
| 82 | - **Output**: Saved to Notes/Stockpick/Stockpick - [Tanggal].md |
| 83 | - **Features**: |
| 84 | - Volume filtering (V > V_avg20) |
| 85 | - Time sensitivity (last 12 hours) |
| 86 | - Multi-source validation |
| 87 | - High conviction setup identification |
| 88 | - Sector diversification |
| 89 | |
| 90 | ## Integration Notes |
| 91 | This approach integrates multiple Hermes skills: |
| 92 | - **crawl4ai** (primary): Self-hosted web scraping via Crawl4AI (free, unlimited) |
| 93 | - **firecrawl** (fallback): For web crawling when Crawl4AI unavailable |
| 94 | - **web_search**: DuckDuckGo search via Crawl4AI wrapper |
| 95 | - **obi-notes**: For saving results to Obsidian vault |
| 96 | - **cronjob**: For scheduling automated execution |
| 97 | - **skill-documentation**: For maintaining skill documentation |
| 98 | |
| 99 | ## Prerequisites |
| 100 | - Crawl4AI: installed in Hermes venv (`pip install crawl4ai` + `playwright install chromium`) |
| 101 | - Python 3.11 (Hermes venv): `~/.hermes/hermes-agent/venv/bin/python3.11` |
| 102 | - Firecrawl (optional fallback): API key for when Crawl4AI fails |
| 103 | - Access to financial data sources |
| 104 | - Obsidian vault configured for saving analysis results |
| 105 | |
| 106 | ## Customization Options |
| 107 | This workflow can be adapted for: |
| 108 | - Different financial instruments (crypto, forex, commodities) |
| 109 | - Different analysis techniques (technical, fundamental, sentiment) |
| 110 | - Different schedules (intraday, daily, weekly, monthly) |
| 111 | - Different output formats (alerts, reports, watchlists) |
| 112 | - Diffe |