Trust & Moderation Guide
Comprehensive guide to trust levels and content moderation in itemID Community Networks — how trust is calculated, content reporting workflows, and moderation tools for community managers.
Trust & Moderation
Community Networks use a layered system of trust levels and content moderation to keep communities safe while empowering active members. This guide covers both systems in detail.
Trust Levels
Every community member has a trust level that determines their daily alert limit. The level is computed automatically based on account activity and moderation history.
| Level | Requirements | Daily Alert Limit | Badge |
|---|---|---|---|
| New | Default for all accounts | 2 | No badge |
| Verified | Account ≥ 30 days old, ≥ 3 registered items, no upheld reports | 5 | Green shield |
| Scout | Same as Verified (higher limits reserved for future criteria) | 10 | Indigo shield |
How Trust Is Calculated
The TrustService evaluates a user's trust level each time it is needed. The following factors are checked in order:
- Upheld reports: If any content report against the user has been upheld, the level is immediately set to New
- Account age: The account must be at least the configured number of days old (default: 30)
- Registered items: The user must have registered at least the configured number of items (default: 3)
Thresholds are configurable via the config/community.php trust section.
Content Moderation
Content submitted to Community Networks passes through a two-layer moderation system before being published.
Automated Checks
- Keyword blocklist: Content is checked against a configurable list of blocked words (case-insensitive). Matches are rejected immediately with a reason.
- AI moderation (optional): When enabled, content is sent to the OpenAI Moderation API for classification. Flagged content is rejected with details on the category. The system fails open — if the API is unavailable, content passes.
Community Reporting
Any authenticated community member can report alerts or sightings that they believe violate community guidelines. Reports include:
- A reason (spam, harassment, misleading, inappropriate, privacy violation, or other)
- Optional details to provide context
Moderator Review
Community moderators (organisation managers with the Moderator role) can review reports through the management dashboard:
- Navigate to Management → Moderation
- Review pending reports with reporter details, reason, and context
- Choose to uphold (confirm the violation) or dismiss the report
- Optionally add resolution notes
When a report is upheld:
- The content creator is notified via email
- The creator's trust level is recalculated (typically reduced to New)
- The report is marked as upheld with the reviewer's notes
Report Reasons
| Reason | Description |
|---|---|
| Spam | Unsolicited, repetitive, or irrelevant content |
| Harassment | Threatening, abusive, or intimidating behaviour |
| Misleading | False or intentionally deceptive information |
| Inappropriate | Content that does not belong in the community context |
| Privacy violation | Exposes someone's personal information without consent |
| Other | Concerns that do not fit the predefined categories |
Configuration
All moderation and trust settings are managed in config/community.php:
- Blocklist: Comma-separated words via the
COMMUNITY_MODERATION_BLOCKLISTenvironment variable - AI moderation: Enable/disable via the
COMMUNITY_MODERATION_LLMenvironment variable - Trust thresholds: Minimum items and days for each level
- Alert limits: Daily alert limits per trust level