This directory provides simple, runnable examples demonstrating the multi-source normalization, deterministic comparison, feed formatting, and tool dispatch capabilities of data2dsl.
| Number | Example | Sources / Scope | Description |
|---|---|---|---|
| 01 | 01-markdown-github-comparison |
Markdown vs GitHub Commits | Factual claim verification comparing documented commit metrics against GitHub API observations. |
| 02 | 02-oql-telemetry-verification |
OQL Spec vs Sensor Telemetry | Hardware-in-the-loop (HIL) verification comparing declared operating specs against sensor logs. |
| 03 | 03-doctor-diagnostic-feed |
Comparison Bundle $\to$ Doctor Agent | Transforming discrepancies into a prioritized diagnostic-profile/v1 for triage agents. |
| 04 | 04-koru-remediation-feed |
Comparison Bundle $\to$ Koru Feed | Generating machine-actionable remediation-intent/v1 payloads for closed-loop self-healing. |
| 05 | 05-mcp-tool-dispatch |
JSON-RPC 2.0 / MCP | Invoking data2dsl_compare over Model Context Protocol (MCP). |
| 06 | 06-closed-loop-self-healing |
Subactor Envelope & Closed Loop | Complete 5-stage closed loop (DETECT -> PLAN -> EXECUTE -> VERIFY -> HEAL) with envelope validation. |
| 07 | 07-sumd-table-comparison |
SUMD Tables vs Telemetry | Factual extraction from Structured Unified Markdown Document tables and deterministic comparison. |
| 08 | 08-batch-multi-query |
Multi-Query Batch & Markdown | Batch evaluation of multiple queries against observation pools with clean ratio aggregation and Markdown report formatting. |
# Example 01: Compare two observations
python src/data2dsl_cli.py compare -l obs_left.json -r obs_right.json -q query.json
# Example 03: Export Doctor Agent Diagnostic Profile
python src/data2dsl_cli.py feed-doctor -b comparison_bundle.json -o diagnostic_profile.json
# Example 04: Export Koru Remediation Intent
python src/data2dsl_cli.py feed-koru -b comparison_bundle.json -o remediation_intent.json