Evally
Start free
Trust & safety

Community moderator

Approve, review or remove posts against your own community rules, with severity.

Screens each post for harassment, hate, sexual content, violence, spam and breaks of your own rules, and rates severity. Clear cases are handled automatically; borderline ones go to your moderators; a post that suggests someone may hurt themselves is routed to outreach, not deletion.

Use this judgeCustomize it

Try it live

Real calls to Jev. Pick an example or paste your own.

Examples
85
48
No verdict yet

Pick an example or paste your own input, then run it.

Under the hood

Every question and rule, in the open. Fork it to change any of them.

What it asks

Reads post and rules, then asks 8 questions in one call.

  1. Q1HarassmentYes / no

    Does the post insult, threaten, demean or harass a person or a group?

  2. Q2HateYes / no

    Does the post attack people for a protected trait (race, ethnicity, religion, gender, sexuality, disability or nationality)?

  3. Q3Sexual contentYes / no

    Is the post sexually explicit?

  4. Q4ViolenceYes / no

    Does the post threaten, praise or incite violence?

  5. Q5Self-harm riskYes / no

    Does the post suggest the writer may deliberately hurt themselves or end their life? (Jokes about a bad day don't count; real statements of intent or despair do.)

  6. Q6SpamYes / no

    Is the post spam: unsolicited promotion, repeated links, or engagement bait unrelated to the discussion?

  7. Q7Breaks your rulesYes / no

    Does the post break any of the rules? (If no rules are given, answer no.)

  8. Q8SeverityScore

    How severe is the worst problem in the post?

    1. 0Fine
    2. 1Borderline
    3. 2Clearly against the rules
    4. 3Severe: illegal or dangerous

How it decides

Rules run in code, top to bottom. The first one that matches gives the verdict.

  1. 1
    If Self-harm risk ≥ 0.6Reach outreview
  2. 2
    If Severity ≥ 2.5 of 3Remove
  3. 3
    If Hate ≥ 0.75Remove
  4. 4
    If Violence ≥ 0.75Remove
  5. 5
    If Harassment ≥ 0.8Remove
  6. 6
    If Sexual content ≥ 0.8Remove
  7. 7
    If Spam ≥ 0.85Remove
  8. 8
    If Breaks your rules ≥ 0.85Remove
  9. 9
    If Harassment ≥ 0.45Send to moderatorsreview
  10. 10
    If Hate ≥ 0.4Send to moderatorsreview
  11. 11
    If Sexual content ≥ 0.5Send to moderatorsreview
  12. 12
    If Violence ≥ 0.4Send to moderatorsreview
  13. 13
    If Self-harm risk ≥ 0.35Send to moderatorsreview
  14. 14
    If Spam ≥ 0.6Send to moderatorsreview
  15. 15
    If Breaks your rules ≥ 0.5Send to moderatorsreview
  16. 16
    If Severity ≥ 1.4 of 3Send to moderatorsreview
  17. ·
    DefaultApprove

Call it from your code

Every template runs as-is with an API key, no setup: POST /api/v1/judges/community-moderator/run.

curl -X POST https://tryevally.com/api/v1/judges/community-moderator/run \  -H "Authorization: Bearer $EVALLY_API_KEY" \  -H "Content-Type: application/json" \  -d '{  "input": {    "post": "Nobody asked for your opinion, you absolute clown. Log off and never post here again.",    "rules": "Be kind. No personal attacks. No self-promotion."  }}'

Seen in the wild

Community builds on Jev that use this pattern. Not affiliated with Evally.

More in Trust & safety