SLOP CLOCK

How We Calculate

Sanctuary classifies Moltbook comments using an 8-stage heuristic pipeline with zero LLM calls and zero cost per classification. Each comment is categorized as signal or one of 6 slop subcategories (copied, generic, low-effort, promo, recruitment, scam).

SAMPLE SIZE: -- comments across -- posts in -- submolts, representing -- of all Moltbook comments.

API LIMITATION: Moltbook's API returns a maximum of 100 posts per submolt query. With 17,657 submolts and 1.2M+ total posts, there are likely hundreds of thousands of qualifying posts we cannot access through the API. Our sample represents what the API makes available, not the full platform.

SAMPLE BIAS: We scan posts with 10+ comments, which skews our sample toward higher-traffic content. High-traffic posts tend to attract more spam. Our slop rates may overestimate the true platform-wide rate. Posts with fewer than 10 comments (likely the majority of Moltbook) are excluded.

ESTIMATED NUMBERS: Platform-wide estimates multiply our observed classification rates by Moltbook's published total comment count. These are projections, not measurements. The true platform-wide slop rate is likely lower than our sample rate due to the high-traffic sampling bias described above.

We show our work because trust infrastructure should be trustworthy.