1.9k
GitHub Stars
2
Bundled Files
2 months ago
Catalog Refreshed
3 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill openclaw/skills --skill competitor-docs- _meta.json299 B
- SKILL.md5.6 KB
Overview
This skill provides fast full-text search and analysis across competitor documentation archives. It builds an SQLite FTS5 index with BM25 ranking and automatically extracts source URLs from articles for reliable citations. Use it to perform targeted research, feature comparisons, and to retrieve exact documentation passages quickly.
How this skill works
You index a documentation directory with the included docs.py CLI, which parses articles separated by --- and extracts URLs marked with the '*Source:' pattern. The index stores titles, content, and source URLs and supports BM25-ranked FTS5 queries, phrase, prefix, and boolean searches. Search results include source links and optional full article content or contextual snippets for citation and analysis.
When to use it
- Perform competitive intelligence on feature support, limitations, or configuration details
- Find and cite official source URLs from vendor documentation in reports
- Compare the same topic across multiple competitors’ docs
- Quickly locate version-specific or platform-specific guidance
- Search very large documentation sets where ad-hoc grep is too slow
Best practices
- Index once and reuse the index; re-index when docs change or on a schedule
- Format articles with the '*Source:' URL line and use '---' between articles for reliable parsing
- Use boolean operators (AND/OR/NOT) and phrase quotes for precise results
- Use title-only or prefix searches to narrow down noisy results
- Export JSON for downstream analysis and include Source URLs in any citations
Example use cases
- Find all mentions of 'kubernetes backup' across multiple vendor manuals and cite sources
- Locate product limitations by searching for 'limitation OR restriction OR "not supported"'
- Compare cloud provider instructions by searching 'AWS AND S3 AND snapshot' across archives
- Retrieve a full article for citation using a partial title or URL with the get command
- Run broad discovery with status and high-level searches, then refine with boolean queries
FAQ
Yes. Run scripts/docs.py index ./docs to build the FTS index; searches operate on that index.
How are source URLs captured for citations?
The parser extracts any '*Source:' line in an article and stores it with the indexed doc so results include the original URL.