Update to PostEverywhere

{
  "description": "Review how social media posts performed with PostEverywhere analytics. Use when the user asks how their posts did, what performed best, which platform works, or wants a weekly or monthly social media report.",
  "included_files": [],
  "name": "review-post-performance",
  "skill_md_contents": "---\nname: review-post-performance\ndescription: Review how social media posts performed with PostEverywhere analytics. Use when the user asks how their posts did, what performed best, which platform works, or wants a weekly or monthly social media report.\n---\n\n# Review post performance\n\nGive the user a short, honest read of how their posts performed, and one or two concrete next steps.\n\n## Steps\n\n1. **Pick the period.** Default to the last 30 days. Call `get_analytics_summary` with `period` (`week`, `month`, or `all`), or with `from` and `to` dates.\n2. **Find the posts.** Call `list_posts_advanced` with `status: \"published\"` and `published_after` for the same period, sorted by `published_at`.\n3. **Look closer where it matters.** For the best and worst few posts, call `get_post_results` to see each platform's outcome and link.\n4. **Write the report.** Keep it short:\n   - Totals for the period: posts published and engagement, from the summary.\n   - The best posts and what they have in common (format, topic, platform, time of day).\n   - Anything that failed to publish, and whether it needs fixing (see `fix-failed-posts`).\n   - One or two suggestions the user can act on this week.\n\n## Be accurate\n\n- Some platforms don't share every number (for example, some don't report views for all account types). Say \"not available\" rather than zero.\n- Don't compare the user's numbers with other people's or invent benchmarks.\n- A small number of posts can't show a trend. Say so when the sample is small.\n"
}