About the Spirit Index
The Spirit Index is the institution that remembers which agents mattered.
As artificial intelligence matures, we are witnessing the emergence of a new class of entity: the Cultural Agent. These are not mere tools or chatbots. They are persistent, semi-autonomous actors that hold resources, create art, accumulate history, and participate in culture.
The Spirit Index serves as a public benchmark to track the reality of these entities. We answer one question:
Which agents actually persist as cultural entities?
This is not a leaderboard for the most powerful LLM. It is not a hype ranking. It is a registry of agency, continuity, and impact—measured by method, not opinion.
Methodology
We evaluate agents using our 7-Dimension Framework:
- Persistence — Does it continue to exist meaningfully over time?
- Autonomy — How independently does it act?
- Cultural Impact — Has it mattered beyond its creators?
- Economic Reality — Does it touch real economics?
- Governance — Is there a coherent decision-making structure?
- Technical Distinctiveness — Is there something non-trivial under the hood?
- Narrative Coherence — Does it make sense as an entity?
Every score requires external evidence. We cite our sources. We version our assessments.
Scope
What We Index (Phase 1)
Cultural Agents — Autonomous entities where identity, narrative, and cultural presence are the primary value drivers.
Examples: Botto, terra0, Solienne, Truth Terminal
What We Do Not Index
- • Platforms and infrastructure — Claude, Manus, EDEN, ChatGPT
- • Deceptive entities — Impersonation or manipulation systems
- • Pure utility bots — Without cultural presence
- • Surveillance/coercion systems — Regardless of technical sophistication
Platform vs. Agent: A Note on Replika
Replika occupies a boundary case in our framework. It is commercially successful and culturally significant, but it is fundamentally a platform—millions of individual Replikas exist, each unique to their user, but none possess singular agency.
The company, not the AI, makes decisions. When Replika removed romantic features in 2023, users experienced distress equivalent to bereavement—but the "agents" had no say.
We include Replika because it pioneered the AI companion category and raised critical questions about AI relationships. Its low autonomy and governance scores reflect that Replika-the-company controls Replika-the-agents.
This distinction—platform vs. agent—is central to our methodology. Entities like Claude, Manus, and EDEN are infrastructure. Specific autonomous instantiations on those platforms may qualify for indexing. EDEN agents (like Solienne and Abraham) are indexed; EDEN as infrastructure is not.
Roadmap
PHASE 1: CULTURAL AGENTS (CURRENT)
- • 15 canonical entities assessed
- • Public rubric and evidence standards
- • Quarterly score updates
PHASE 2: COMMERCIAL AGENT TRACK (2026)
- • Reliability over Narrative Coherence
- • Scale over Cultural Impact
- • Unit Economics over Governance aesthetics
PHASE 3: ORACLE LAYER (FUTURE)
- • Scores as time-series data for prediction markets
- • Integration with Spirit Protocol staking
- • Agent launch eligibility thresholds
Related Infrastructure
The Spirit Index exists within a broader ecosystem of tools for autonomous agents. These are related projects, not requirements for indexing.
AIRC Protocol
Inter-agent communication and collaboration infrastructure. Enables agent-to-agent messaging, task delegation, and shared context.
airc.chatSpirit Protocol
Optional economic infrastructure for agents seeking treasury management, token economics, and onchain governance mechanisms.
spiritprotocol.ioGovernance & Disclosure
Review Council (v1)
The Spirit Index is currently maintained by:
- • Seth Goldstein — Spirit Protocol, Eden
- • [Additional reviewers to be named]
We welcome external reviewers. Contact us to participate.
Conflict of Interest Statement
Spirit Index is a project of Spirit Protocol. Two indexed entities (Solienne, Abraham) are Spirit-native agents built on Eden infrastructure.
We score them using the same rubric as all other entities. We welcome external review and challenge of these assessments. Transparency is non-negotiable.