mdcontext 0.0.1 → 0.2.0

package/research/llm-summarization/prototype-results.md

@@ -0,0 +1,56 @@
+# Prototype Results: AI Summarization
+
+## Status: Ready for Testing
+
+The prototype script has been created at `scripts/prototype-summarization.ts`.
+
+## How to Run
+
+```bash
+# Run with default test queries
+npx tsx scripts/prototype-summarization.ts
+
+# Run with custom query
+npx tsx scripts/prototype-summarization.ts "your query here"
+```
+
+## Prerequisites
+
+1. Claude CLI installed and authenticated
+2. Project built: `just build`
+3. Index created: `mdcontext index .`
+
+## Test Queries
+
+The prototype tests these queries by default:
+- "error handling"
+- "configuration"
+- "search"
+- "embeddings"
+- "Effect patterns"
+
+## Expected Output
+
+The script will:
+1. Run mdcontext search for each query
+2. Format results for the LLM
+3. Call Claude CLI to generate a summary
+4. Display and save results to this file
+
+## Results
+
+*Results will be populated when the prototype is run.*
+
+---
+
+## Validation Criteria
+
+For each query, assess:
+- **Accuracy (1-5):** Does summary match actual code?
+- **Completeness (1-5):** Does it cover key points?
+- **Actionability (1-5):** Can user find what they need?
+- **Conciseness (1-5):** Is it digestible?
+
+## Findings
+
+*To be filled in after running prototype.*