Analyze content performance data to identify statistically significant winning patterns.
npx skills add https://github.com/cgallic/kai-cmo-harness --skill content-retro
Please help me install this Agent Skill. Skill: content-retro Source: https://github.com/cgallic/kai-cmo-harness/tree/main/content-retro Command: npx skills add https://github.com/cgallic/kai-cmo-harness --skill content-retro
Analyze content performance data to extract winner patterns and automatically update learning defaults, ensuring future content improves without manual tuning.
Run /content-retro after publishing content to refresh learned defaults and improve future planning.
High-intent search queries and answers about installing and using this skill.
To identify winning content patterns, you analyze published performance data to surface statistically significant trends in hooks, formats, personas, word counts, and publish days. This process computes win rates and updates learned defaults to guide future briefs.
A content feedback loop analyzes performance data to extract winner patterns and automatically update learned defaults. This mechanism ensures future content planning improves continuously without requiring manual tuning of your strategy.
You analyze content performance data by loading graded pieces into a retro process that computes win rates and statistical significance across variables like hook types and timing. It then stores a pattern archive to close the loop.
Yes, you can automate learning defaults across a content portfolio by scaling pattern detection across graded published pieces. The system computes win rates and auto-updates defaults to continuously improve future content outcomes.
You should run a content performance retro immediately after publishing content. This timing refreshes learned defaults from the latest performance logs, ensuring future planning and briefs are guided by the most recent statistically significant patterns.
Pattern detection analyzes content variables including hook types, formats, personas, word counts, and publish days. By computing win rates and statistical significance for these elements, the system identifies winning combinations to guide future briefs.
robit-manDetect performance regressions by comparing latency and throughput benchmarks across software versions.
nextcEvaluates extracted patterns and routes them to appropriate folders.
cgallicGenerate an 18-field content brief from format, site, and keyword inputs.
zmreExecute Fabric patterns natively in Claude to process content and return structured outputs.
