Render LLM responses as Markdown with Rich

Add first LLM API practice

next working step

working state

Initial commit

Add eager_student agent: homework_doing variant of angry_teacher

Add eager_student agent: homework_doing variant of angry_teacher

Refactor angry teacher agent and enhance functionality

Implement angry teacher agent and update project configuration

+ mcp

Implement perplexity agent and update dependencies

fix: env var validation, SSL error handling, clear error messages

chore: add .env.example and .gitignore

+presentation

docs: update README v2.5 changelog and observability sample

fix: v2.5 — critical bugs + idempotency + feedback loop

feat: v2.4 — per-letter timeout, observability, confluence tree, pre-filter fix

docs: update README with agentic architecture, ReAct loop, and full tool descriptions

fix: clean up imports across all agent files

feat: add LLM reranking to knowledge base search

feat: add real escalation email sending

feat: refactor to agentic ReAct loop with tool calling