searching-web-and-files

A WIP collection for finding, retrieving, researching, verifying, and synthesizing information across the web and local files.

Contents

• skills/ — the original research collection plus the searching-web-and-files router.
• library/ — a broad, deduplicated dump mined from the local skill library.
  • methodology/ — iterative research, synthesis, verification, and subagent orchestration.
  • web-tools/ — search engines, crawlers, scrapers, browsers, and extraction APIs.
  • file-code-retrieval/ — filesystem/codebase exploration, repository search, RAG, vector search, and context retrieval.
  • specialized-research/ — academic, market, customer, competitive, company, trend, patent, and grant research.
• inventory/selected-library.json — provenance, hashes, duplicate paths, and copy destinations for imported items.
• inventory/broad-candidates.json — the wider machine-ranked candidate pool retained for later mining.

The initial seed was copied from:

/Users/jordan/skill-workspace/resources/staging/registry/files/research

The local scan examined 6,647 case-insensitive SKILL.md files and retained 281 distinct related items after lexical scoring, independent agent audits, and duplicate collapsing. Imported content remains unchanged; entry files are renamed to SOURCE.md so the dump stays disclosed reference material rather than becoming hundreds of automatically loaded skills.

Start with skills/searching-web-and-files/SKILL.md. Its catalog links into the full dump.