# Explore Ontology

> Explore ontology questions, grounded concepts, evidence, actual task results and evolution outcomes in a read-only interactive version explorer.

- Skill: `ruc-datalab/explore-ontology` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add ruc-datalab/explore-ontology`
- Raw SKILL.md: https://api.skillmd.com/api/skills/ruc-datalab/explore-ontology/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Productivity
- Author: ruc-datalab (https://skillmd.com/u/ruc-datalab)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/ruc-datalab/explore-ontology

---


# Explore Ontology

Render all ontology versions as one standalone, offline, single-file interactive
HTML explorer (Results & questions / Content / Schema / Tool views), with in-page version switching
and side-by-side comparison across all three layers. Difference highlighting can
be toggled without changing the compared versions. The operation is strictly
read-only.

## Workflow

1. Resolve the workspace input: default is the current project's
   `.evoontology/`; use an explicit path when the user provides one. The Core
   accepts an exact workspace, a `.evoontology` container, or a project root.
   It uses the requested version to discover one matching nested workspace at
   any depth. If multiple workspaces match, report the candidates and ask for
   the exact path instead of guessing.
2. Resolve the initially shown version: default `active` (the version referenced
   by `active.json`). An explicit `ontology_vN` changes only the initial page
   selection and never changes `active.json`; every available version is embedded.
3. Call the `evo-semantic` MCP tool `visualize_ontology` (the single rendering
   entry point), passing the absolute workspace input and optionally `version`
   and `open_browser`. Do not run
   `python -m evoontology.visualization`.

4. In Codex desktop call visualize_ontology with presentation:"codex" and
   open_browser:false. The tool returns html_path and a loopback browser_url.
   Open browser_url with the available Codex open_in_codex tool using
   target:{type:"browser",url:browser_url}, placement:"right". Reuse the preview
   tab when available. The MCP tool does not open an external browser in this mode.
   If the app browser tool is unavailable, return the URL and HTML path explicitly;
   do not claim the page opened. For an explicitly requested external browser use
   presentation:"external". No browser is opened twice.

## Boundaries

- Do not implement HTML generation inside this skill; all rendering lives in
  EvoOntology Core (`evoontology.visualization`).
- Never modify Build, Evolve, Runtime, `active.json`, `versions/`, or any
  other ontology/evolution state.
- Errors stay explicit: workspace not initialized, no active version, the
  requested version does not exist, or multiple nested workspaces match.
  Broken references only produce warnings; graph objects are never fabricated.
- Content edges follow the semantic model: solid Semantic Relations connect
  Terms, while dotted Structural References attach Mapping, Constraint, and
  Evidence records according to schema reference rules.

The Results & questions page distinguishes semantic coverage, observed execution
and independent evaluation. Click a question to highlight its associated semantic
objects. Terminal run summaries show provenance-labelled metrics and public task
replays; no reserved validation cases are exposed. Missing reports or executions
remain explicit empty states. Never fabricate examples to fill the page.