Guided, not magical
Prompts teach the capability surface while DVQR engines remain the source of deterministic evidence and interpretation.
v0.15.9 · Discoverability & Guided Investigation
Understand Dataverse. Know what to ask. Follow the evidence.
DV Quick Run is an extension-owned Local MCP server and investigation workbench for VS Code. Start from one of 94 guided prompts, ask in plain English, inspect deterministic Dataverse evidence, and continue into deeper managed investigation when needed.
Quick Start → Prompt → Evidence → Suggested next step. v0.15.9 adds 94 curated prompts, 69 Free and 25 Pro, while the live MCP surface now exposes 24 Free tools plus 32 Pro-only tools—56 total with Pro.
Prompt Library · 94 guided prompts
Product philosophy
Discover the capability. Preserve the evidence. Bound the conclusion.
Evidence-first
Metadata confidence, runtime observations, unavailable evidence, and historical snapshots remain distinct.
Progressive disclosure
Quick Starts orient newcomers; categories, search, filters, and follow-ups support deeper workflows.
Human authority
DVQR narrows investigation and execution choices. People still decide what is true, safe, and appropriate.
Prompt Library · v0.15.9
Go from “what can I ask?” to a guided DVQR journey.
DV Quick Run now includes a category-driven Prompt Library with 94 curated user-intent prompts, Quick Starts, search, Free/Pro filtering, parameterised prompt rendering, and deterministic suggested next prompts.
Browse 69 Free and 25 Pro prompts across six categories without learning MCP tool names first.
Start from common outcomes such as finding tables, understanding relationships, checking DVQR Score, discovering Custom APIs, or investigating an issue.
Follow deterministic next-prompt links from discovery into evidence-backed querying, traversal, verification, and managed investigation.
Prompt Library · v0.15.9
Operational Profile through Free MCP
Ask for the live Operational Profile and canonical DVQR Score.
The new Free MCP Operational Profile capability uses the same canonical evidence and scoring path as the GUI, with calibrated-score transparency, semantic guardrails, and explicit live-result authority over historical snapshots.
Inspect weighted evidence contributors and the calibrated 0–100 DVQR Score without hidden residual contributors.
Keep relationship fanout, plug-in steps, orchestration participation, ownership, packaging, and unknown evidence semantically bounded.
Use the score as operational-density orientation—not a risk, health, performance, security, or root-cause score.
Operational Profile through Free MCP
Managed Investigation foundation · Pro
Continue from guidance into persisted, evidence-backed investigation.
v0.15.9 builds on the v0.15.8 managed-investigation protocol: persistent intent, deterministic strategy, bounded evidence acquisition, readiness, Mini RCA checkpoints, and handoff remain the deeper Pro workflow.
Persist investigation identity, evidence provenance, readiness, checkpoint freshness, and continuation state.
Acquire exactly bounded evidence through deterministic actions rather than conversational reconstruction.
Keep observed participation and runtime reachability separate from mechanism, causality, and business authority.
Managed Investigation foundation · Pro
Free foundation. Pro acceleration.
Start with guided discovery. Go deeper when the investigation needs persistence.
Free includes 24 MCP tools for Dataverse understanding, Operational Profile + DVQR Score, bounded queries, metadata and relationship intelligence, and basic eligible Custom API workflows. Pro adds 32 Pro-only MCP tools—56 total—including persisted managed investigations, target-aware runtime evidence, readiness, Mini RCA, comparison, timeline, reporting, and evidence continuity.