Use when authoring, reviewing, or maintaining an `/llms.txt` file or per-page markdown mirrors per the llmstxt.org specification. Triggers on edits to `llms.txt` / `llms-full.txt`, on prompts about LLM-friendly site content, providing markdown versions of pages (`.md` suffix), generating expanded context files (`llms-ctx.txt`), or making documentation accessible to LLMs — even when the user doesn't say 'llms.txt'.
How this skill is triggered — by the user, by Claude, or both
Slash command
/xonovex-skill-llmstxt:llmstxt-guideThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Author and maintain a project's `/llms.txt` file (and the per-page markdown mirrors it points to), so LLMs can ingest the site at inference time within a bounded context window.
Author and maintain a project's /llms.txt file (and the per-page markdown mirrors it points to), so LLMs can ingest the site at inference time within a bounded context window.
A markdown file served at /llms.txt (or a subpath) that gives an LLM a curated, concise map of a site's most useful content. Complements — does not replace — robots.txt (which is about crawler permissions) and sitemap.xml (which is exhaustive and HTML-oriented).
The file must contain these elements in this order:
Only the H1 is required. The other sections are optional but follow this ordering when present.
## Section name
- [Link title](https://example.com/page.md): Optional notes
- [Another link](https://example.com/other.md)
Each list item is:
[name](url) (required): and brief notesAn H2 section literally titled ## Optional designates secondary resources that can be skipped when a shorter context is needed. Processors use this distinction to produce two variants of expanded context.
For every HTML page worth giving to LLMs, serve a clean markdown version at the same URL with .md appended:
https://example.com/docs/intro.html → https://example.com/docs/intro.html.mdhttps://example.com/docs/intro → https://example.com/docs/intro.mdhttps://example.com/docs/intro/ → https://example.com/docs/intro/index.html.mdhttps://example.com/ → https://example.com/index.html.mdThe llms.txt file links to these .md mirrors, not the HTML originals.
When a site provides llms.txt + linked .md mirrors, a processing tool (e.g. llms_txt2ctx) can produce expanded single-file context bundles:
llms-ctx.txt — expansion without the ## Optional URLs (shorter context)llms-ctx-full.txt — expansion including the ## Optional URLs (full context)Some projects also publish llms-full.txt as a single-file dump of everything (e.g. https://hono.dev/llms-full.txt) — this is a common convention even though the spec primarily defines llms.txt.
.md must actually exist at the URLllms.txt is a guide, not a sitemap; aggregate-too-large content is the whole point of the format.md mirror is the #1 authoring mistake — LLMs receive raw HTML with nav chrome, breaking the whole purposeindex.html.md, not just .mdllms-full.txt (single dump) is a popular sibling convention but not part of the spec — don't confuse it with the spec's llms-ctx-full.txt (which is a processing output)## Optional is a literal section title with semantic meaning to processors — don't rename it ("Extra", "Bonus", "Additional") if you want the short-context variant to work/llms.txt from scratch for a project or site.md mirrors that /llms.txt links to (URL conventions, content rules)/llms.txt implementations to model your own on/llms.txt + mirrors into a single-file context bundle (llms-ctx.txt, llms-ctx-full.txt)Guides completion of development work by verifying tests, detecting environment, and presenting structured options for merge, PR, or cleanup.
Guides creation and editing of skills using test-driven development with pressure scenarios and subagents to verify agent compliance.
Dispatches multiple subagents concurrently for independent tasks without shared state. Use when facing 2+ unrelated failures or subsystems that can be investigated in parallel.
npx claudepluginhub xonovex/platform --plugin xonovex-skill-llmstxt