kb-re-structure

Organize free-form dumped notes into a canonical knowledge base structure with attribution.

1|1|Updated Jun 21, 2026
One-click install
npx skills add https://github.com/Navdeepgambhir9023/nav --skill kb-re-structure-navdeepgambhir9023
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kb-re-structure
Source: https://github.com/Navdeepgambhir9023/nav/tree/main/skills/kb-re-structure
Command: npx skills add https://github.com/Navdeepgambhir9023/nav --skill kb-re-structure-navdeepgambhir9023

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Free-form notes dumped into folders like dump/, notes/, or inbox/ are hard for agents and teammates to use. This Skill reorganizes that raw content into a canonical kb/ structure (product, architecture, rules, glossary, decisions) without losing or silently overwriting anything. ## Core Features & Use Cases - Classification into canonical KB files: Each dumped file or section is classified into product.md, architecture.md, rules.md, glossary.md, or decisions.md based on content signals, with a confidence-rated table presented for approval before any writes. - Lossless archiving: Source dumps are moved to kb/.archive/<timestamp>-<folder>/ with an ARCHIVE_INDEX.md recording every file, its size, and its KB destination. - Attributed, idempotent merges: Imported content is appended under a dated ## Imported YYYY-MM-DD from <filename> marker, never overwriting existing curated sections, so re-runs merge safely. - Use Case: A founder pastes weeks of scattered product notes into a dump/ folder; the skill inventories them, proposes a classification table, archives the originals, and merges each chunk into the right KB file with full provenance. ## Quick Start Ask the agent to reorganize the notes in your dump folder into the kb knowledge base structure, for example: "Reorganize everything in my notes/ folder into the kb structure."

Frequently Asked Questions about kb-re-structure

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I organize scattered notes into a project knowledge base?

Point the skill at your dump folder (or let it scan for dump/, notes/, inbox/, or kb-dump/). It inventories every file, proposes a classification into product, architecture, rules, glossary, and decisions files, and merges content only after you approve the table.

How does the skill decide which KB file each note belongs to?

Classification uses content signals: users and features go to product.md, components and data flow to architecture.md, must/never rules to rules.md, term definitions to glossary.md, and tradeoff or ADR language to decisions.md. Ambiguous chunks are flagged for your review rather than guessed.

Will reorganizing notes overwrite my existing knowledge base content?

No. Existing curated sections are never overwritten. New content is appended below them under a dated 'Imported YYYY-MM-DD from <filename>' marker, and the original dump files are preserved in kb/.archive/ with an index.

What happens to the original dump files after reorganization?

Source files are moved (or copied, if you prefer) into kb/.archive/<timestamp>-<folder>/ with filenames preserved. An ARCHIVE_INDEX.md lists every archived file, its line count, format, and KB destination so provenance stays traceable.

Can I re-run the restructure skill on new notes later?

Yes, the skill is idempotent. Re-running merges new dumped content under fresh dated attribution markers without disturbing previously imported or curated sections, and each run creates its own timestamped archive folder.