LL-030: Discord Community Plan (Server Structure and Governance)

Server Architecture Overview

Loki's Lab's Discord server operates as a community hub for collaboration, benchmark discussion, and contribution coordination. The infrastructure is designed to separate public discourse from operational verification while maintaining a clear hierarchy of authority.

Category and Channel Structure

The server contains four primary categories containing specific channels dedicated to function:

1. Announcements (Category)
2. #welcome-and-rules: Mandatory read, outlines community guidelines and website authority clauses.
3. #server-status: Official updates regarding server maintenance or migration events.
4. Discussion (Category)
5. #general: Open chat for non-operational community engagement.
6. #hardware-and-models: Specific topic discussion on supported hardware and model architectures.
7. #workflow-talk: Dialogue regarding testing procedures and research methods.
8. Test Requests (Category)
9. #test-request-intake: Public channel for submitting request ideas; uses a reaction-based voting system before moderator selection.
10. #in-progress-tests: Live logs for current official tests only.
11. Contributor Operations (Category)
12. #help-desk: Dedicated support for contributors needing technical or procedural help.
13. #submission-reviews: Review channel where work is vetted prior to publication.

Role Model and Hierarchy

Role permissions are assigned based on contribution verification status rather than tenure:

• @Member: Default role with read access to general channels. Cannot post in test or submission channels without moderator approval.
• @Contributor-Auto: Automated assignment for low-risk data submissions following standard protocols (verified email, valid JSON structure). Limited write permissions in specific categories.
• @Verified: Manual assignment only. Indicates successful completion of benchmark verification steps. Has access to high-level discussion channels.
• @Trusted: Manual assignment by senior staff or maintainers. Grants administrative oversight permissions within community management tools.

Authority and Data Integrity Policy

Discord is a communication platform, not a repository of truth. All users must adhere to the following rule:

• The Website Authority Rule: The website remains the single source of authority for all benchmark records. No benchmark record exists only in Discord. Discord posts are illustrative and secondary references to official data. If a discrepancy exists between Discord logs and the website, the website takes precedence.

Permanent Invite Policy Note

To ensure long-term security and prevent spam or hijacking through link flooding:

• Public invite links are managed centrally and are permanent but rate-limited by the system bot.
• Direct server invites cannot be generated by users or moderators using standard Discord features without approval from Operations.
• All external sharing of the community is subject to moderation review prior to activation.

LL-033: Discord Role Automation + Manual Trust Gates

Automated vs. Manual Assignment Logic

To prevent trust manipulation and ensure data integrity, role assignment utilizes a hybrid model of automation for low-risk tasks and manual oversight for high-trust status.

Low-Risk Automation Scope

Only the @Contributor-Auto role is managed via Discord bot automation scripts. This automation triggers only when specific criteria are met:

1. Validated Identity: User has passed email verification and account security checks.
2. Data Integrity: A valid JSON submission file exists in the system with matching hashes.
3. Publication Status: The associated record is publicly listed on the website.

Rule: Users do not qualify for automation roles without passing these three gates. If a user lacks an automated role tag but still possesses content, it indicates a Manual Trust status.

High-Trust Manual Oversight

The @Verified and @Trusted roles are strictly manual administrative assignments. No bot automation grants access to these permissions. These roles require direct human verification