Social Source Calibration

This tile helps agents interpret socially noisy web sources without overfitting to the vibe of the crowd.

It contains four source-specific skills:

• moltbook-source-calibration — for Moltbook research, where signal can hide inside a lot of spam, noise, and novelty-seeking posts.
• hacker-news-source-calibration — for Hacker News research, where experienced technical skepticism can surface real structural concerns but also over-weight cynicism.
• reddit-source-calibration — for Reddit research, where large-sample user sentiment and recurring pain can be useful, but expertise and evidence quality vary wildly by subreddit and comment thread.
• product-hunt-source-calibration — for Product Hunt research, where launch-day momentum, upbeat comments, and leaderboard position can be informative about positioning and early reaction but should not be mistaken for durable product truth.

Use these skills when research involves summarizing, distilling, or weighting evidence from those communities. They are not fact-checking skills on their own; they are calibration layers for interpreting crowd-generated material that has already been gathered by the surrounding research workflow.