Scan and chunk markdown/txt content for classification into PAI taxonomy destinations.
npx skills add https://github.com/davdunc/pai-framework --skill migrate-davdunc
Please help me install this Agent Skill. Skill: Migrate Source: https://github.com/davdunc/pai-framework/tree/main/skills/Migrate Command: npx skills add https://github.com/davdunc/pai-framework --skill migrate-davdunc
Migrates existing content into the correct PAI destination taxonomy so your TELOS, preferences, and knowledge become usable inside your Claude Code–based assistant.
Run the migrate flow on a directory of notes by asking: migrate my notes from the path /path/to/content.
High-intent search queries and answers about installing and using this skill.
Migrating existing notes into a personal AI system involves scanning and chunking markdown files, then classifying each chunk for routing into structured TELOS, preferences, and knowledge categories. A confidence-scored approval loop ensures correct placement before committing.
Importing CLAUDE.md and .cursorrules files into a personal AI memory system involves scanning the exported instruction and rules content, chunking it, and routing it to the correct destination taxonomy. Provenance tags are stamped on each commit to preserve the source origin.
Migrating content from prior PAI USER, TELOS, and MEMORY folders preserves provenance by stamping each approved commit with source origin data. The migration process scans existing folder structures and routes chunks into the correct destination categories with confidence-scored approvals.
Chunk classification handles low-confidence or UNCLEAR content by forcing careful review during the migration approval loop. High-confidence chunks are auto-approved, medium-confidence items are routed for confirmation, and UNCLEAR chunks require manual verification before committing to the destination taxonomy.
The migration workflow does not require manual approval for every chunk. Confidence thresholds auto-approve high-confidence chunks, route medium-confidence items for confirmation, and force careful review only for low-confidence or UNCLEAR content.
Migrating plain text and markdown files to a PAI taxonomy is limited by the confidence-scoring accuracy of chunk classification. UNCLEAR or low-confidence chunks require manual review, and all commits must pass through the provenance-tagged approval loop before routing to TELOS or knowledge outputs.
starlink-awakenClassify and route external content into PAI taxonomy destinations.
starlink-awakenOrchestrate phased conversational interviews across PAI context files to validate TELOS.
Kastalien-ResearchQuery, add, and rank codebase learnings using memory CLI tools.
wasintohRoute user requests to appropriate agents based on intent and context.
