find hidden workflows

The problem

Users often spend hours repeating the same prompts, context-setting, and manual workflows without realizing their routines could be automated. Identifying these patterns manually by scrolling through months of chat history is tedious and results in missed opportunities for optimization.

What it does

• Analyzes accessible chat history to detect semantic routines and repeated task patterns across professional and personal domains.
• Distinguishes between recurring interests and actual repeatable workflows with similar inputs, judgments, and outcomes.
• Audits existing skills to recommend concrete improvements rather than creating redundant duplicates.
• Scores opportunities based on frequency, context burden, and reusability to prioritize high-value automations.
• Generates structured briefs for the skill-creator tool to turn identified patterns into functional agent skills.

Why this beats prompting it yourself

Identifying workflows requires cross-referencing hundreds of disparate conversations for semantic similarity, which is nearly impossible for a human to do objectively. This skill uses a multi-pass search strategy and a weighted scoring ledger to find non-obvious patterns that a simple "what should I automate?" prompt would miss.

Use cases

• Finding "hidden" administrative tasks like weekly reporting or email drafting that consume recurring manual effort.
• Identifying complex decision-making processes that require consistent logic and specific context.
• Turning repeated personal research habits into structured, reusable planning tools.
• Optimizing existing custom instructions that are failing to capture new variations in your work style.

Known limitations

Analysis is restricted to accessible chat history and may miss conversations if retrieval tools are unavailable. It identifies patterns but requires user approval before invoking the skill creation process.