Add issue log for maintenance follow-ups

ISSUES.md

@@ -0,0 +1,21 @@
+# Proposed Maintenance Tasks
+
+## Typo Fix
+- **File:** `1-Introduction_Agents/2-structured.py`
+- **Issue:** The Pydantic model for calendar events is named `CalenderEvent`, which is a misspelling of "Calendar" and can be confusing in tutorials or imports.
+- **Task:** Rename the class to `CalendarEvent` and update its references for clarity.
+
+## Bug Fix
+- **File:** `1-Introduction_Agents/4-retrieval.py`
+- **Issue:** `search_kb` opens `b"Introduction_Agents\\know-db.json"`, mixing a bytes path with Windows-style separators and the wrong directory name. This crashes on POSIX systems because the file cannot be found.
+- **Task:** Build the knowledge-base path with `Path`/`os.path` using the correct `1-Introduction_Agents/know-db.json` location.
+
+## Documentation/Comment Discrepancy
+- **File:** `4-smolagent/NewsAgent.py`
+- **Issue:** The footer claims the app is "Powered by Groq LLama 3.1 8B Instant" while the code instantiates the `deepseek-r1-distill-llama-70b` model, leading to misleading documentation.
+- **Task:** Update the caption (or model selection) so the displayed model name matches the actual configuration.
+
+## Test Improvement
+- **File:** `4-smolagent/NewsAgent.py`
+- **Issue:** `create_analysis_prompt` builds a detailed multi-section prompt, but there is no automated coverage ensuring the required sections stay intact when the function changes.
+- **Task:** Add a unit test (e.g., `tests/test_news_agent.py`) that asserts each numbered section appears in the generated prompt for a sample topic, preventing regressions in prompt structure.