Files
Nathan Schneider bda868cb45 Implement LLM-driven governance architecture with structured memory
This commit completes the transition to a pure LLM-driven agentic
governance system with no hard-coded governance logic.

Core Architecture Changes:
- Add structured memory system (memory.py) for tracking governance processes
- Add LLM tools (tools.py) for deterministic operations (math, dates, random)
- Add audit trail system (audit.py) for human-readable decision explanations
- Add LLM-driven agent (agent_refactored.py) that interprets constitution

Documentation:
- Add ARCHITECTURE.md describing process-centric design
- Add ARCHITECTURE_EXAMPLE.md with complete workflow walkthrough
- Update README.md to reflect current LLM-driven architecture
- Simplify constitution.md to benevolent dictator model for testing

Templates:
- Add 8 governance templates (petition, consensus, do-ocracy, jury, etc.)
- Add 8 dispute resolution templates
- All templates work with generic process-based architecture

Key Design Principles:
- "Process" is central abstraction (not "proposal")
- No hard-coded process types or thresholds
- LLM interprets constitution to understand governance rules
- Tools ensure correctness for calculations
- Complete auditability with reasoning and citations

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
2026-02-08 14:24:23 -07:00

6.8 KiB

Peer-to-Peer Dispute Resolution

A self-facilitated process where participants work together directly to resolve disputes

This dispute resolution protocol can be integrated into any governance constitution as an article on conflict resolution.


Article: Dispute Resolution

Section 1: Principles and Values

Core Values This community emphasizes:

  • Open dialogue and direct communication
  • Mutual respect between all parties
  • Collaborative problem-solving
  • Focus on solutions rather than blame
  • Voluntary participation with encouragement

Community Standards

  • Standards maintained in shared digital and physical formats
  • Reviewed annually by the community
  • Accessible to all members
  • Bot maintains current version

Section 2: Initiating Dispute Resolution

When to Use Members should initiate dispute resolution for:

  • Interpersonal conflicts between members
  • Disagreements about community practices
  • Misunderstandings requiring clarification
  • Relationship repair needs

How to Initiate

  1. Submit incident report to @govbot
  2. Document: parties involved, events, evidence, desired outcomes
  3. Available in electronic or paper format
  4. Bot notifies all parties and provides process guidance

Voluntary Participation

  • Participation is voluntary but strongly encouraged as first step
  • Treated as good faith effort to resolve conflicts
  • If party declines, coordinator reaches out to understand concerns
  • Alternative paths available if unsuitable

Section 3: The Peer-to-Peer Process

Self-Facilitation Participants manage the process themselves:

  • No third-party facilitator required
  • Bot provides guidance on communication techniques
  • Participants choose meeting location
  • Scheduling arranged mutually

Ground Rules All participants agree to:

  • Take turns speaking without interruption
  • Use "I" statements about personal experience
  • Ask clarifying questions to understand better
  • Summarize understanding to confirm accuracy
  • Focus on solutions rather than dwelling on problems
  • Respect confidentiality of the process

Meeting Structure Typical process (60-90 minutes):

  1. Each person shares their perspective
  2. Ask clarifying questions
  3. Identify common ground and differences
  4. Joint brainstorming of potential solutions
  5. Agree on specific actions or outcomes

Section 4: Assessing the Dispute

Joint Assessment Participants work together to identify:

  • Specific issues that need addressing
  • How each person has been impacted
  • Relevant community values at stake
  • What resolution would look like
  • Requirements for moving forward

Scope and Jurisdiction This process is suitable for:

  • Most interpersonal conflicts between members
  • Communication breakdowns
  • Disagreements about behavior or actions
  • Relationship tensions

Escalation Criteria Must escalate to higher level if dispute involves:

  • Illegal activity
  • Safety risks to individuals or community
  • Harassment or serious code of conduct violations
  • Power imbalances requiring facilitation support

Section 5: Deliberation and Problem-Solving

Discussion Process

  • Open conversation about the situation
  • Each person's needs and concerns heard
  • Clarifying questions encouraged
  • Joint exploration of options
  • Creative brainstorming of solutions

Adding Support

  • Initially involves direct parties only
  • If deadlocked, may invite one mutually trusted person
  • Support person helps facilitate, doesn't decide
  • Keeps process peer-to-peer focused

Reaching Conclusion Process concludes when:

  • Participants feel issues thoroughly explored
  • Ready to make decisions about resolution
  • Clear about agreements and next steps
  • Or agree to escalate to facilitated process

Section 6: Resolution Outcomes

Types of Resolution Successful peer-to-peer process may result in:

  • Clarifications clearing up misunderstandings
  • Apologies for harm caused
  • Behavioral agreements for future interactions
  • Restoration of harm (returning items, making amends)
  • Agreed boundaries for future relationship
  • Recognition of different perspectives

Documenting Agreements

  • Parties can document agreements if desired
  • Submit to @govbot for record-keeping
  • Not required but recommended for accountability
  • Bot sends reminders if follow-up scheduled

Mutual Agreement Required

  • Both parties must agree to any resolution
  • No imposed outcomes in peer-to-peer process
  • Partial agreements acceptable
  • Can agree to disagree on some points

Section 7: When Resolution Doesn't Work

Escalation Path If peer-to-peer doesn't resolve the issue:

  1. Acknowledge that escalation is needed
  2. Refer to trained mediator pool via @govbot
  3. Mediators trained in more formal processes
  4. Professional referral connections available if needed

No Penalty for Escalation

  • Escalation is normal, not a failure
  • Shows good faith effort was made
  • Some disputes need more structured support
  • Community values trying direct resolution first

Section 8: Follow-Up and Accountability

Checking Agreement Follow-up available when:

  • Agreement isn't working as expected
  • Circumstances have changed significantly
  • One party requests check-in
  • Scheduled follow-up time arrives

Requesting Follow-Up

  1. Either party submits written request to @govbot
  2. Request focuses on specific agreement issues
  3. Initiates new conversation
  4. May adjust agreements as needed

Learning and Improvement

  • Bot tracks anonymized patterns (not individual details)
  • Quarterly summaries help community learn
  • Information limited to involved parties
  • Success patterns shared to help others

Section 9: Information and Privacy

Confidentiality

  • Details limited to parties directly involved
  • Not shared publicly without consent
  • Bot maintains secure records
  • Anonymized data only for community learning

Access to Information

  • Parties have access to their own case documentation
  • Quarterly anonymized summaries shared with community
  • Statistics help improve process
  • Individual privacy protected

Record Retention

  • Bot maintains dispute resolution records
  • Available to parties for reference
  • Supports accountability to agreements
  • Helps track patterns needing community attention

Implementation Notes for Bot

When facilitating peer-to-peer dispute resolution:

  1. Make process accessible - Provide clear, simple guidance
  2. Support self-facilitation - Offer communication tips without taking over
  3. Respect autonomy - Let parties control their process
  4. Track agreements - Help with follow-up and accountability
  5. Enable escalation - Make it easy to get more support when needed
  6. Protect privacy - Keep details confidential
  7. Learn from patterns - Use anonymized data to improve community

This process works best when community culture supports direct communication, members feel empowered to handle conflicts, and higher-level support is available when needed.