Extract CLI corrections and review patterns from Claude Code session JSONL files.
npx skills add https://github.com/vinhnxv/rune --skill learn-vinhnxv
Please help me install this Agent Skill. Skill: learn Source: https://github.com/vinhnxv/rune/tree/main/plugins/rune/skills/learn Command: npx skills add https://github.com/vinhnxv/rune --skill learn-vinhnxv
Extracts CLI corrections, review recurrences, and meta-QA patterns from recent Claude Code session history to improve Rune Echoes memory and long-term workflow quality.
Invoke /rune:learn to scan recent session history, run detectors, and persist patterns to Rune Echoes memory.
High-intent search queries and answers about installing and using this skill.
You can extract CLI corrections and review recurrences from session JSONL files by running targeted detectors over recent Claude Code session history to surface high-confidence patterns. The process scans JSONL data and persists findings into the echoes store.
Capturing meta-QA patterns from Claude Code session history involves running meta-qa detectors over recent sessions to identify recurring quality insights. The detected patterns are then persisted to a local echoes store for future workflow automation.
Yes, you can preview detected learning signals before persisting them by using a dry-run mode. This mode scans session JSONL files and runs detectors to show results without writing the high-confidence patterns into the echoes store.
Yes, real-time pattern detection is supported via a watch mode that continuously monitors session JSONL files. This allows detectors to run automatically and capture CLI corrections and review recurrences as new session data is generated.
You can run targeted detectors for specific learning patterns by selecting from available options like cli, review, arc, hook, and meta-qa. These detectors operate on session JSONL files and TOME findings to extract and persist high-confidence results.
Persisting CLI corrections and review recurrences automates the capture of learning signals from session JSONL files, improving long-term workflow quality. Storing high-confidence patterns in an echoes store enables future automation and prevents recurring manual reviews.
marcusExtract Claude Code local data schemas from JSONL sessions and stats-cache.json.
jakechen1Capture user corrections and record them as durable learning rules.
Fl2vioConfigure AI agents to persist and retrieve memory across sessions using AgentDB.
jonmastProcess tabular data with the qsv CLI for filtering and cleaning.
