session-skill-discovery

You can identify reusable skill candidates from session transcript logs by parsing JSONL session data with Python scripts to extract procedural patterns across multiple tickets and sessions, then applying quality gates to validate the results.

Session analysis for skill discovery is the process of examining post-completion Claude Code session transcript logs to detect repetitive procedural patterns that can be extracted and validated into reusable skill candidates.

Yes, you need Python 3 along with standard libraries like json and re to parse JSONL transcript logs and execute the pattern recognition scripts required for identifying reusable skill candidates.

You validate extracted skill candidates by applying built-in quality gates that check the identified procedural patterns against existing skill documentation and criteria to ensure only high-quality candidates are proposed.

Yes, you can analyze cross-ticket patterns by running the Python scripts against multiple session transcript logs simultaneously, allowing you to identify procedural patterns that span multiple projects and tickets.

Using regex and JSON parsing for skill discovery is limited to structured post-completion analysis of transcript logs, meaning it cannot perform real-time session monitoring and relies entirely on the procedural patterns captured within the JSONL data.