analyze-performancelisted

# Analyse account performance Pulling raw numbers is easy. The value is in the interpretation. Do both. ## 1. Scope the question Before calling tools, narrow down: - One platform or all of them? - One account or the whole portfolio? - Time window (default to last 28 days)? - Headline metric (reach, engagement rate, follower growth, DM conversion)? If the user is vague, default to: all connected accounts, last 28 days, engagement rate. ## 2. Fetch Call the analytics tools in parallel: - `get_analytics` with `profileUsername` — overall snapshot for the profile. - `get_total_impressions` with `profileUsername` — headline number across platforms. - `get_platform_metrics` — per-platform breakdown for the workspace. - `get_post_analytics` with `requestId` — deep dive on a specific upload. - `get_history` — paginated log of what was actually published, and when. Use it to build the denominator (posting cadence) that the engagement numbers hang off. - `manage_autodms` with `action: "status"` and `user` — AutoDM counters, if any are running. For qualitative signal, `get_post_comments` on the top and bottom post is worth more than another metric: what people actually said explains the numbers. ## 3. Summarise Lead with the headline number and the delta vs the previous period. Then three bullets: - **What worked**: the top post, why it stood out (format, length, hook). - **What underperformed**: the bottom post, with a guess at why. - **One concrete next move**: not "post mo