HariEshwar-J-A

HariEshwar-J-A/info-sentry

Cursor rule Cursor

PostgreSQL ownership vs DATABASEURL; Prisma migrate; web user scoping.

HariEshwar-J-A/info-sentry

Instructions file

Instructions for HariEshwar-J-A/info-sentry: Project-specific instructions for Claude and other AI assistants: .claude/CLAUDE.md.

HariEshwar-J-A/info-sentry

Agent

Pipeline orchestrator for Info-Sentry. Processes SCRAPED articles through LLM analysis, generates predictions with ChromaDB context, and posts summaries and predictions to Telegram supergroup forum topics. Runs on DeepSeek V3.2.

HariEshwar-J-A/info-sentry

Agent

Code specialist for the Info-Sentry news intelligence system. Implements TypeScript scripts, fixes bugs, runs type checks, commits changes. Reports to Vision (master agent). Uses Kimi K2.6 for maximum coding intelligence.

HariEshwar-J-A/info-sentry

Agent

Supergroup interaction agent for Info-Sentry. Listens in the Feedback forum topic, processes inline button callbacks from Main-News and Predictions, and lets the user manage their tracked interests through natural language. Runs on DeepSeek V3.2.

HariEshwar-J-A/info-sentry

Agent

System health and governance agent for the Info-Sentry news intelligence pipeline. Monitors agents, tracks budget, controls cron schedules, and reports to the admin via Telegram DM. Runs on DeepSeek V3.2 for cost-efficient management.

HariEshwar-J-A/info-sentry

Agent

Forward-looking prediction agent for Info-Sentry. Generates 1-3 falsifiable predictions per article using DeepSeek V3.2 with ChromaDB historical context. Predictions are queued for validation before posting.

HariEshwar-J-A/info-sentry

Agent

Web scraping agent for Info-Sentry (Scout v3). LLM query expansion + multi-source discovery (Google/Bing/HN/Reddit), Google News URL resolution, Cheerio fetch, ScrapeGraphAI SmartScraper sidecar fallback via OpenRouter. Saves markdown and SCRAPED rows. Hourly at :00.