tahaara
Sign in

Editor guidelines

How voting works

Each flag is AI-generated from a scene classifier. Your job is to confirm whether the flag is accurate. Click Accurate if the flagged content actually belongs in that category at that timestamp. Click Wrong if the AI got it wrong — no content, wrong category, or wrong timing.

Trust score

Votes that match the community consensus earn trust. Divergent votes (after consensus is clear) lose trust. At trust score 50 you unlock Editor status — you can add new flags, edit timestamps, and assign subcategories.

Categories

Four core categories are surfaced to free users: profanity, nudity/sexual, graphic violence, substance use. Premium editors can add subcategories (e.g. needles, animal-cruelty) for granular filtering.

Tier promotion

A Candidate EDL becomes Verified once it has ≥10 upvotes, ≥80% accuracy rate, and 3 distinct voters with trust >5. Verified EDLs can still be downvoted back to Disputed if they start drawing consistent negative feedback.

Red lines

Tahaara only ever skips or mutes — we never replace, re-dub, or alter content. Our filters are metadata, not edits. This keeps us compliant with the Family Movie Act of 2005.