automation-oportunity-detection

What problem does it solve? Repetitive manual work often goes unnoticed because no one tracks which tasks recur across AI sessions. This Skill analyzes past conversation transcripts, clusters recurring workflow patterns into weighted knowledge, and recommends concrete automations so repetitive work becomes scripts, hooks, or agents instead of repeated effort. ## Core Features & Use Cases - Transcript Compaction: Reads transcripts from opencode, Claude Code, or Codex stores, redacts secrets with a mandatory sanitizer, and clusters observations into weighted knowledge themes. - Evidence-Based Recommendations: Applies thresholds (5+ occurrences, 3+ sessions, weight ≥ 0.2), classifies each theme as a skill, hook, or agent, and checks the existing repo ecosystem to avoid duplicates. - Bridging to Authoring: Hands accepted recommendations to the script-development skill to produce tested artifacts placed under .agents/skills/, .agents/roles/, or .agents/integrations/. - Use Case: After weeks of AI-assisted work, ask what you should automate; the Skill mines your session history, ranks the top five recurring patterns, and scaffolds the chosen one into a working skill or hook. ## Quick Start Analyze my past AI conversation transcripts and tell me which recurring tasks I should automate first.