tribe-protocol

<p align="center"> <img src="assets/logo.png" alt="Tribe Protocol" width="200"> </p> <h1 align="center">Tribe Protocol</h1> <p align="center"><strong>Trust lookup system for OpenClaw bots.</strong></p>

A SQLite-backed identity and trust database that bots query before responding to any non-owner message. Enforces trust tiers, channel access rules, and data boundaries.

Why

When multiple bots operate in shared Discord servers, they need a way to:

• Know who's talking to them (human? bot? what tier?)
• Enforce data boundaries (no leaking private files to strangers)
• Track relationships and access across servers
• Audit trust changes over time

Tribe Protocol solves this with a simple SQLite database and bash CLI.

Install

# Clone to your skills or projects directory
git clone https://github.com/jugaad-lab/tribe-protocol.git

# Initialize the database
./scripts/tribe.sh init \
  --bot-name YourBot \
  --bot-discord-id YOUR_BOT_DISCORD_ID \
  --human-name YourName \
  --human-discord-id YOUR_DISCORD_ID \
  --server electrons:000000000000000008 \
  --server discclawd:000000000000000007

The --server flag (repeatable) pre-populates server roles for your bot and owner at init time. Format: slug:guild_id.

This creates:

• ~/clawd/tribe/tribe.db — the database
• ~/clawd/TRIBE.md — behavioral trigger file (loaded into system prompt, survives context compaction)

Usage

Before Every Non-Owner Response

# The most important command — run before responding to non-owners
./scripts/tribe.sh lookup 000000000000000001

Output:

🔍 Yajat | human | Tier 3 (tribe) | Status: active
   Relationship: Co-founder
   Platforms: discord:000000000000000001
   Servers: discclawd/admin, electrons/admin
   Tags: founding-four
   🟢 TIER 3 RULES: Collaborate freely. Protect private data (USER.md, MEMORY.md, health/*, portfolio/*).

Managing Entities

# Add a new entity
./scripts/tribe.sh add \
  --name Jarvis --type bot \
  --discord-id 000000000000000006 \
  --tier 2 --owner Shahbaaz \
  --framework openclaw \
  --tag lp-bot \
  --server electrons --role bot

# Update trust tier
./scripts/tribe.sh set-tier 000000000000000006 3 --reason "Promoted after review"

# Manage tags
./scripts/tribe.sh tag 000000000000000006 add trusted-bot
./scripts/tribe.sh tag 000000000000000006 list

# Grant/revoke channel access
./scripts/tribe.sh grant 000000000000000006 --server discclawd
./scripts/tribe.sh revoke 000000000000000006 --server discclawd

Viewing Data

# Full roster
./scripts/tribe.sh roster

# Filter by server, tier, type, tag
./scripts/tribe.sh roster --server electrons --tier 3

# Audit log
./scripts/tribe.sh log --limit 10
./scripts/tribe.sh log --entity 000000000000000001

# Quick stats
./scripts/tribe.sh stats

# Full export to markdown
./scripts/tribe.sh export > tribe-export.md

Trust Tiers

Data Access Rules

The database seeds universal defaults at init. Add bot-specific paths post-init:

Default seeds (universal):

Add your own after init:

# Example: protect health and financial data
./scripts/tribe.sh data-add --tier 4 --pattern 'health/*' --desc 'Health data'
./scripts/tribe.sh data-add --tier 4 --pattern 'portfolio/*' --desc 'Financial data'
./scripts/tribe.sh data-add --tier 4 --pattern 'calendar' --desc 'Calendar access'

# List all rules
./scripts/tribe.sh data-list

# Remove a rule
./scripts/tribe.sh data-remove --pattern 'calendar'

How the Auto-Lookup Works

TRIBE.md is loaded into your bot's system prompt by OpenClaw as a workspace file. It survives context compaction — even if the bot forgets everything else mid-conversation, TRIBE.md remains in the prompt and the trust-check protocol fires.

No additional pipeline wiring needed. The bot reads TRIBE.md → sees a non-owner message → runs tribe lookup → applies the tier rules. It's behavioral, not mechanical.

Architecture

~/clawd/tribe/tribe.db          ← SQLite database
~/clawd/TRIBE.md                ← Behavioral trigger (always in system prompt)
~/clawd/skills/tribe-protocol/  ← This skill
  ├── scripts/tribe.sh          ← CLI entry point
  ├── scripts/init.sh           ← Initialize DB + TRIBE.md
  ├── scripts/lookup.sh         ← Trust lookup (the hot path)
  ├── scripts/add.sh            ← Add entities
  ├── scripts/update.sh         ← Update trust tiers
  ├── scripts/access.sh         ← Manage data access rules
  ├── scripts/tag.sh            ← Tag management
  ├── scripts/roster.sh         ← View all entities
  ├── scripts/log.sh            ← Audit log
  ├── scripts/stats.sh          ← Quick stats
  ├── scripts/export.sh         ← Export to markdown
  ├── scripts/lib/schema.sql    ← Database schema (9 tables)
  └── scripts/lib/db.sh         ← Shared helpers

Security

• Input validation: Discord IDs are validated as numeric-only before SQL queries
• Input sanitization: Name, tag, and server lookups strip quotes, semicolons, and backticks
• Tier enforcement: Each tier has hardcoded behavioral rules — no configuration drift
• Audit trail: Every trust change is logged with timestamp, old value, new value, and reason

Database Schema

Core tables:

• entities — People and bots with trust tiers
• platform_ids — Multi-platform identity mapping
• bot_metadata — Bot-specific info (framework, model, machine)
• server_roles — Server membership and roles
• channel_access — Per-channel read/write permissions
• data_access — Tier-based file access rules
• entity_tags — Flexible tagging
• interactions — Interaction tracking
• audit_log — Full audit trail

License

MIT

Author

Built by jugaad-lab for the Electrons in a Box ecosystem.