Andrey Oz
Community@ozand
Andrey Oz maintains a knowledge base skill suite for capturing error lessons, semantic search, and documentation indexing across developer session histories.
Agent Skills by Andrey Oz
Showing 8 vetted skills indexed across 1 GitHub repositories.
ozandkb-harvest
Ingest HTML, PDF, JSON, and TXT documentation into the QMD semantic docs index.
ozandkb-search
Searches lessons, session logs, and docs using QMD hybrid semantic and BM25 retrieval.
ozandkb-mine
Mine agent session histories to detect recurring errors and suggest lesson capture.
ozandkb-enrich
Generates cross-links, backlinks, citations, and indexes for knowledge base lessons.
ozandkb-visualize
Generate an interactive HTML graph visualization of knowledge base lessons.
ozandkb-capture
Creates a new error lesson file and registers it in the knowledge base index.
ozandkb-sync
Synchronize a local knowledge base with a remote git repository via pull and push.
ozandkb-lookup
Matches error text against error_signatures in index.yaml to retrieve knowledge base fixes.
Frequently Asked Questions About Andrey Oz
FAQPage SchemaWhat tasks can I perform with Andrey Oz's kb skill suite?▼
You can capture error-resolution lessons, look up fixes via index.yaml signatures, ingest HTML/PDF/JSON/TXT docs into a QMD semantic index, run hybrid semantic search across lessons and session logs, mine histories for recurring errors, enrich lessons with cross-links and citations, visualize the Knowledge Base as an HTML graph, and sync with a remote repository.
Who is the target audience for these Knowledge Base skills?▼
Developers and engineers who maintain persistent, searchable records of resolved errors and lessons learned. The suite suits individuals or teams wanting to mine agent session logs, index external documentation, and retrieve past fixes through semantic search rather than re-solving known problems.
How do the kb skills work together in practice?▼
A typical flow: kb-capture records a lesson after fixing an error, kb-enrich adds cross-links, citations, and backlinks, kb-harvest ingests external docs into the QMD index, kb-search and kb-lookup retrieve fixes, kb-mine surfaces recurring errors from logs, kb-visualize refreshes the HTML graph, and kb-sync pushes changes to the remote repository.
What data sources and formats do these skills support?▼
The suite ingests external documentation in HTML, PDF, JSON, and TXT formats via kb-harvest. It also processes curated lessons, agent session histories, and session logs for search and mining, with error signatures stored in index.yaml for fast lookup.
What are the prerequisites for using the kb skills?▼
You need an existing local Knowledge Base structure with an index.yaml file for error signatures, plus a remote repository if you intend to use kb-sync for synchronization. Session histories and logs are required for kb-mine to identify recurring error patterns.