Finding what makes your content perform
Two tools turn your posting history into concrete guidance.
Content patterns
get_content_patterns analyzes your post history and surfaces what actually correlates with performance:
- Hook styles: how your different opening lines perform
- Post lengths: whether short or long posts do better for you
- CTA impact: what including a call to action does to your numbers
- Posting times: when your posts land best
It is computed entirely from history Crispy already has, with zero additional LinkedIn calls. Scope the analysis with an optional days parameter (minimum 7) to focus on recent posts.
It needs at least 10 posts to produce meaningful results and returns empty below that. This is deliberate: patterns extracted from a handful of posts would be noise dressed up as insight.
Comment reach
get_comment_reach answers a different question: is time spent commenting on other people's posts actually driving visibility? Over a chosen window (default 30 days) it returns which posts you commented on, impression correlation, and a daily breakdown. It returns empty when no comment activity is cached for the selected period.
Putting them together
A monthly loop that works: run get_content_patterns, take the two strongest findings (say, question hooks and mid-length posts), and encode them into your posting style guide with set_writing_style so every future agent draft follows them automatically. Then run get_comment_reach to decide whether commenting deserves a bigger or smaller share of your daily engagement budget.
Both tools read cached data only, so they are free to run and safe to wire into a recurring agent workflow.