Examples

See It In
Action

From scanning your codebase to querying it through your AI agent. Copy, paste, and run.

Scan a Codebase

Build the knowledge graph. Fast structural scan — no model loaded.

bash
# Fast structural scan (FTS5 search ready in seconds)
devscriptor scan ./my-project

# Only Python files, exclude tests
devscriptor scan ./my-project -i py -e tests/,*_test.py

# Re-scan only changed files
devscriptor scan ./my-project --incremental

# Scan + embed inline (needs ~4 GB free RAM)
devscriptor scan ./my-project --embed

Embed & Search

Add semantic search, then query by keyword or meaning.

bash
# Generate vectors once (idempotent, model unloads after)
devscriptor embed

# Keyword / symbol search
devscriptor search "getUserById"
devscriptor search "connection pool" -t fulltext

# Semantic similarity search (requires embed)
devscriptor find-similar "database connection pooling"

Analyze & Inspect

Static analysis, call graphs, and a health check.

bash
# Dead-code and full analysis
devscriptor analyze ./my-project -k dead-code
devscriptor analyze ./my-project -k all -o json

# Call graph from an entry point (Mermaid output)
devscriptor call-graph --entity main --format mermaid --max-depth 3

# Database stats and health check
devscriptor status
devscriptor doctor