Scoring, Filtering, and Statuses
ThreadLens scores and filters findings so a project does not become a pile of raw links.
Scoring
Section titled “Scoring”Posts are scored for signals such as pain, relevance, frustration, solution seeking, and workaround language. Google results are analyzed for relevance, intent, and opportunity.
Filtering
Section titled “Filtering”Filtering removes low-signal or promotional results before they consume review time. Deduplication keeps repeated URLs and posts from appearing as separate findings when they represent the same signal.
Statuses
Section titled “Statuses”Use statuses to track review progress:
| Status | Meaning |
|---|---|
new | Found but not reviewed. |
starred | High-signal and worth returning to. |
excluded | Not useful for the current project. |
drafted | Used for a draft response or follow-up. |
commented | Already acted on. |
Keep statuses factual so reports and future reviews stay useful.
Filtering is visibility, not scoring
Section titled “Filtering is visibility, not scoring”Scout can hide spam, bot-like, low-quality-account, and likely AI-generated findings from default review queues. This filtering does not change post_score, final_score, Google relevance scores, draft fields, or review statuses.
Filtered findings are retained for the self-host owner under Filtered findings. Use Restore visibility to recover only one item, or Restore and trust… to recover the item and create a project-scoped trust record for the displayed source or exact system-generated filter signature. Trust records are allowlist overrides; they do not boost scores.
Use Re-filter selected from the normal post list to check selected visible posts again. Use Re-check selected from Filtered findings to apply new trust records to already-filtered rows.