Mission 04: Create Onchain Ontology for Intuition Ecosystem Apps

[jeremie-olivier]

Board item: Create the onchain Ontology for each Intuition ecosystem apps using ontology.intuition.box
Status: Ideas | Priority: P0
Reward: $500 USDC + iBox token
Target: Any builder familiar with Intuition ontology concepts

Context

The Intuition ecosystem has ~15 community apps built on the protocol, hosted under the Intuition-Box GitHub org. Each app uses the protocol differently but none have a formal onchain ontology definition. The goal is to define a full schema (atom types, predicates, valid relationships) for each app using Ontology.

Objective

For each ecosystem app, create a complete ontology definition onchain via Ontology:

• Define the atom types the app uses or introduces
• Define the predicates (relationships) relevant to the app
• Define valid type combinations (which subjects can relate to which objects via which predicates)
• Publish all definitions onchain as atoms and triples

Ecosystem Apps (from Intuition-Box org)

#	App	Repo	Description
1	Ontology	Ontology	Ontology explorer and visualizer
2	Atlas	atlas	TBD
3	Fee Proxy Template	Fee-Proxy-Template	Fee collection proxy for MultiVault
4	Sofia Fee Proxy Dashboard	Sofia-Fee-Proxy-Dashboard	Dashboard for Fee Proxy management
5	Community Directory	community-directory	Community member directory
6	Spread Trust	spread-trust	Trust propagation tool
7	L3 Node	l3-node	L3 node infrastructure
8	RPC	RPC	RPC endpoint tooling
9	MCP Inspector	mcp-inspector	MCP server inspector
10	9atunments	9atunments	TBD
11	AgentId	agentId	Agent identity system
12	Orbyt	orbyt	TBD
13	Intuition Ideation	Intuition-ideation-skill	Ideation workflow skill
14	Graph	graph.intuition.box (?)	Trust graph visualization
15	Values	Values (?)	Organization values tool

Deliverables

• Identify and confirm the full list of ecosystem apps
• For each app, define:
  • Atom types used (e.g., Agent, Skill, Fee, Vault, etc.)
  • Predicates and relationships
  • Valid type combinations (subject-predicate-object rules)
• Publish each app's ontology onchain via Ontology
• Document the ontology for each app (can be in the app's repo or a central doc)

Acceptance Criteria

• Each ecosystem app has its ontology published onchain
• Ontologies are visible and browsable on ontology.intuition.box
• Schema definitions are consistent across apps (shared types/predicates where applicable)

Tool: https://ontology.intuition.box
Org: https://github.com/intuition-box

[Gmin2]

Hey @jeremie-olivier I want to work on this!

[y4motion]

Hey @jeremie-olivier 👋

I'd like to take on Mission 04. I'm already deep in the Ontology codebase — I've submitted PR #11 for Mission 01 with full Intuition Protocol integration (wallet connect, indexer hooks, onchain atom/triple creation).

My plan:

1. Audit all 15 ecosystem repos to identify atom types, predicates, and valid relationships
2. Create a unified schema with shared types where applicable (e.g., Agent, Skill, Vault appear across multiple apps)
3. Document each app's ontology with structured JSON-LD definitions
4. Publish onchain via the Ontology tool once Mission 01 integration is deployed

I'll start with the repo analysis immediately and share progress here.

[Kryptoremontier]

Hi @jeremie-olivier — I'd like to take this mission on.

I've been publishing canonical atoms and predicates onchain on Intuition mainnet for a while now. A few of the primitives already live in the ontology.intuition.box glossary came from my work — resolved to, true, false, related to, belongs to (via PR #7, reviewed with @Saulo), plus has agent skill and reported for from AgentScore. So I'm already operating inside this exact vocabulary and the onchain publishing flow (Portal + mainnet, rich atoms with descriptions, publisher-filter hygiene notes).
Approach
For each ecosystem app I'll:

Read the repo to identify the atom types it actually uses/introduces and the predicates that describe its core triples.
Reuse existing glossary predicates wherever they apply (the predicates already onchain — supports, trust, has tag, is associated with, etc.) so schemas stay consistent across apps, and only mint new atoms where there's a genuine gap.
Define valid type combinations (subject → predicate → object rules) per app.
Publish onchain as atoms + triples so they show up in the glossary, and document each app's ontology (per-repo or a central doc — your call on where you'd prefer it).

Reusing the shared layer is the key to the "consistent across apps" acceptance criterion — I'd rather extend the existing vocabulary than fork a parallel one.
Scope
I'll cover all the apps that actually exist in the org. Quick note from auditing the list: a few repos don't resolve (community-directory, l3-node, mcp-inspector, 9atunments) and a few are TBD (atlas, orbyt). I'll confirm the final list as deliverable #1 and flag anything that needs a repo or a description from the team. Realistically that's ~11 apps with content today.
Proposed pilot (this week)
To de-risk before minting the full set, I'll start with Fee Proxy Template and AgentId onchain — two apps I know well (I've built on the FeeProxy pattern and on agent identity). I'll publish their full ontologies, link the glossary entries here so you can see exactly what the deliverable looks like, and adjust the structure based on your feedback before rolling out the rest.
Timeline

Pilot (2 apps onchain + glossary links): ~3-4 days
Remaining apps after structure sign-off: ~2 weeks
Central ontology doc alongside

One question for the team: for the pilot I'll mint on mainnet (full conviction, like PR #7) unless you'd prefer a testnet preview first — happy either way, just let me know which you'd rather review.
Links for context: PR #7 · AgentScore predicate integration guide (the publisher-filter pattern is documented there).

Looking forward to it 🙏