Author: @Othman

Category: Legitimacy & Engagement

Status: Draft

Discussion-to: HSP-006: House of Stake - Mission, Vision & Values

Description: Establishes a core governance memory layer for HoS, enabling transparent decision history, outcomes, and future AI tooling.

Type: Sensing

Abstract

This proposal introduces a Governance Memory System (GMS) for NEAR House of Stake — a structured, data-driven layer designed to document proposal lifecycles, track outcomes, surface informal power dynamics, and synthesize recurring governance patterns. The goal is to increase transparency, strengthen legitimacy, and create long-term institutional memory so future decisions are informed by historical context rather than reinvented each cycle.

Context & Alignment

In MVV update F-MVV-107, the authors noted:

“A DAO Data Strategy is needed… Full adoption of a Governance Memory System (GMS) could be achieved via a separate proposal.”

This proposal directly fulfills that directive.

GMS aligns with:

• Value #5: Transparency & Accountability

• Value #8: Data-Informed Decision-Making

• Value #9: Iteration, Adaptation & Feedback Loops

• The AI Governance Dashboard roadmap (summaries, sentiment, revisions, analysis)

GMS does not replace the dashboard or any existing process. It serves as the missing meta-governance layer that ties everything together.

Situation

The DAO currently lacks:

• lifecycle metadata for decisions

• outcome tracking for passed proposals

• institutional memory for new contributors

• synthesis across cycles

• analysis of recurring governance tensions

• visibility into informal power

• objective assessments of whether governance is improving

This leads to repeated debates, confusion, and decision-making without historical grounding. If we don’t act, governance quality will deteriorate as complexity grows.

Mission

To establish a lightweight, modular Governance Memory System that provides:

• traceability of decisions

• clear feedback loops

• better alignment with MVV values

• improved delegate comprehension

• reduced governance amnesia

• stronger institutional legitimacy

Success will be measured using explicit KPIs including:

• 90% metadata coverage

• 70/50/30 completion rate for ORA reviews

• quarterly GHI publication

• 25% reduction in repeated proposals without referencing precedent

Approach

GMS is implemented through five interconnected layers, each providing a different dimension of governance memory.

Layer 1: Proposal Lifecycle Metadata (PLM)

Structured metadata including:

• authorship

• proposal type & domain

• lifecycle state + timestamps

• revision history

• discussion surfaces

• on-chain linkages

• implementation status

Layer 2: Outcome Review Anchors (ORA)

Structured 30/60/90-day reviews assessing:

• whether outcomes match intent

• unintended consequences

• success metrics

• follow-up recommendations

Layer 3: Informal Power Mapping (IPM)

Lightweight analytics that surface:

• recurring authors

• influence clusters

• sentiment asymmetries

• discourse dominance

Layer 4: Governance Health Index (GHI)

A quarterly diagnostic scoring:

• inclusiveness

• transparency

• feedback mechanisms

• accountability

• coordination infrastructure

Layer 5: Recurring Themes & Frictions (RTF)

Cycle-by-cycle synthesis identifying:

• repeated issues

• unresolved tensions

• thematic clustering

• governance bottlenecks

Each component of GMS will be tracked against measurable KPIs to ensure verifiable progress.

Technical Specification

PLM Fields (JSON Example)

• {

• “proposal_id”: “”,

• “title”: “”,

• “author”: “”,

• “proposal_type”: “”,

• “domain”: “”,

• “lifecycle_state”: “”,

• “created_at”: “”,

• “updated_at”: “”,

• “onchain_link”: “”,

• “implementation_status”: “”

• }

Outcome Anchor Fields

• {

• “review_30d”: “”,

• “review_60d”: “”,

• “review_90d”: “”,

• “kpi_status”: “”,

• “observations”: “”,

• “follow_up”: “”

• }

Discussion Analytics Fields

• {

• “sentiment”: “”,

• “top_themes”:,

• “unanswered_questions”: ,

• “consensus_state”: “”

• }

Data Access

• CSV/JSON exports

• Optional lightweight API in future phases

No contract changes or breaking changes are introduced.

Backwards Compatibility

Fully backwards compatible:

• No changes to governance rules

• No changes to voting models

• No protocol-level modifications

• Existing proposals can be manually backfilled

Scope (Phase 0 Only)

Phase 0 focuses on foundational data capture, not analytics or automation yet:

1. Decision Archive

Canonical record of all proposals

• Full text, metadata, authorship

• Linkage to discussion threads

• Timestamped versioning

2. Participation & Process Metadata

• Who participated?

• When?

• Through what channels?

• In what capacities?

3. Outcome Logging

For each decision:

• Was it implemented?

• What changed?

• What’s the current status?

• What were the downstream effects?

4. Feedback Loop Registry

Every proposal tagged with:

• Inputs

• Decisions

• Outcomes

• Lessons learned (even if preliminary)

5. Format Compatibility for Future AI Agents

• Open standards

• Machine-readable formats (JSON, markdown index files, structured logs)

• Clear interfaces for summarization or analysis

Milestones

Phase 1 (0–1 month): Data Foundations

• PLM field creation

• Lifecycle tracking

• Export functionality

Phase 2 (1–3 months): ORA Pilot

• 30/60/90-day reviews for 3–5 proposals

• Report publication

Phase 3 (4 months): First GHI Report

• Full governance cycle health check

Phase 4 (3–6 months): RTF + IPM Activation

• Recurring issue clustering

• Informal influence mapping

• Synthesis reports

Phase 5 (6+ months): Agent Integration

• Feed GMS data into AI delegates

Budget & Resources

The GMS can launch phase 0 without funding initially. It would be good to see what the contribution guidelines develop into.

There can be an optional Micro-Budget for a Pilot

Used for:

• outcome-review contributors

• metadata stewards

• synthesis writers

• data cleaning and reporting

This accelerates adoption but is not required for Phase 0.

Team & Accountability

Lead: @Othman

Supporting Roles:

• Dashboard maintainers (PLM & export integration)

• Screening Committee (rubric alignment)

• Volunteer reviewers / funded cohort

Reporting:
Monthly updates will be posted in the Governance Forum.

Security Considerations

• No smart contracts touched

• No new trust assumptions

• No escalation of privileges

• All data remains public and auditable

• Risks primarily involve interpretation quality, mitigated by clear structure

Conflict of Interest

No conflicts of interest declared.

Success Metrics (KPIs)

1. PLM Coverage

Target: 90% of proposals have complete metadata
Metric: PLM completeness ratio

2. ORA Completion

Target: 70% (30-day), 50% (60-day), 30% (90-day) completion
Metric: ORA interval completion rate

3. GHI Publication

Target: 1 GHI report per governance cycle
Metric: Binary (complete/incomplete)

4. Discussion Analytics

Target: 80% of proposals include sentiment & themes
Metric: Analytics coverage rate

5. RTF Synthesis

Target: 1 synthesis report per quarter or if a pattern triggers one (enough proposals should have been through the entire governance cycle. A cycle being from the naissance of the proposal to the outcome review)
Metric: Reports delivered (complete/incomplete)

6. Delegate Comprehension

Target: 70% of delegates report increased clarity
Metric: Post-cycle survey results

Copyright

Copyright and related rights waived via CC0

Thank you very much for your initiative and the idea!
At the moment, it’s a bit difficult for me to fully understand what the governance memory system would look like in practice, and how it would differ from the existing tools we already have in HoS, such as the dashboard or the forum. The HoS website and the forum already store the full history of proposals, including authorship, and the forum also shows the revision timeline of each proposal.

Perhaps sharing a few successful examples of similar memory systems in action could help illustrate how this approach differs from what we currently have, and what specific value it would add.

Thank you again for your work and contribution!

GMS is a framework, and in practice it would either live as its own website or as a dedicated page on the HoS site. The dashboard and forum are helpful, but they cover only a small slice of what GMS is meant to do. A lot of ecosystems don’t even have that — NEAR does — but we still lack the structure that turns raw governance history into actual insight.

A simple example: before the inflation reduction proposal, we had the forum but no clear voting procedure or consensus rules. That led to confusion around legitimacy. Another example is the recent inflation proposal. It passed, and since then no follow-up, no outcome analysis, no verification of whether the intended effects actually showed up. We just move on to the next proposal.

GMS closes that loop. It makes sure that in the future, when a similar proposal comes up, we’re not guessing — we have concrete data on what worked, which stakeholders participated, where the consensus process slowed down, and what unintended issues emerged.

Phase 0 is intentionally simple: mostly cleaning and structuring the outputs coming from the dashboard so we can get a reliable picture of governance as a whole. The real value shows up in later phases with the Governance Health Index and the Recurring Themes & Frictions layer, which only work if we first establish clean institutional memory.

Long term, this data becomes fuel for any AI governance tools we build, because it contains structured context the forum and dashboard alone don’t provide. And if helpful, I can share a case study from the Juno ecosystem where I retroactively applied a similar process.

@Othman this is great to see, and I love the build on our Mission and Values!

I can see this Governance Memory System could build into something hugely valuable, from this initial proposed starting point.

Are you able to move it into the Proposals subcategory where all proposals are being shared? Otherwise we can ask someone to help with that

Thanks for the suggestion, just moved it!

Appreciate the proposal, @Othman! Tagging @jacklaing here, this could be a great fit for the product strategy.