Improve docs AI readiness (#59577)

Context This PR makes the Zed docs easier for AI tools, search crawlers, and users to consume without changing the visible docs content. The current production docs are primarily optimized for browser navigation. They do not expose first-class Markdown URLs, an `llms.txt` index, page-level copy affordances, or machine-readable freshness metadata that let users and agents grab clean, current page content. Changes - Generate Markdown copies for docs pages during the mdBook postprocess step, including `/docs/index.md` as an alias for Getting Started. - Generate `/docs/llms.txt` from the mdBook chapter list, grouped by `SUMMARY.md` sections and annotated with page frontmatter descriptions. - Generate `/docs/sitemap.xml` with `<lastmod>` values for every docs page. - Emit machine-readable freshness metadata in HTML via `last-modified` and `article:modified_time` meta tags. - Generate Cloudflare Pages `_redirects` for `.html`, extensionless, and `.md` redirect variants, with channel-aware docs destinations. - Add discovery hints for agents and crawlers: `rel="llms.txt"`, `rel="alternate" type="text/markdown"`, and a short generated `llms.txt` directive in copied Markdown pages. - Update the docs proxy so `Accept: text/markdown`, `/docs.md`, and direct `.md` requests can resolve to the generated Markdown artifacts. - Move primary docs content earlier in the HTML source while preserving the visible layout, so crawlers and agent scorers encounter the article before sidebar chrome. - Move the existing copy-as-Markdown control from the top navigation into the page-title row, using the generated Markdown alternate link as the source of truth. - Split AI-discovery artifact generation out of `docs_preprocessor/src/main.rs` into a focused module. Best Practices Adopted - Use `llms.txt` as a concise navigation index, not a dump of full page content. - Link to absolute, canonical Markdown URLs from `llms.txt`. - Preserve the docs hierarchy in `llms.txt` instead of emitting a flat sitemap-like list. - Include short per-link descriptions from existing metadata rather than inventing summaries. - Keep `llms.txt`, Markdown copies, sitemap data, redirects, and freshness metadata generated from the same mdBook source to avoid drift. - Advertise Markdown alternates with standard HTML metadata and same-origin URLs. - Support both explicit Markdown URLs and content negotiation for clients that prefer Markdown. - Keep browser copy behavior pointed at generated Markdown alternate links instead of duplicating route inference in JavaScript. - Keep the copy-as-Markdown affordance in the page title row without duplicating header chrome controls. Validation - `cargo check -p docs_preprocessor` - `cargo test -p docs_preprocessor` - `./script/clippy -p docs_preprocessor` - `mdbook build ./docs --dest-dir=../target/deploy/docs/` - `node --check docs/theme/plugins.js` - `pnpm dlx prettier@3.5.0 docs/theme/plugins.js --check` - `git diff --check` - Local artifact checks confirmed generated Markdown pages, `llms.txt`, `sitemap.xml` lastmod values, HTML freshness metadata, Markdown alternate links, redirect targets, and preprocessed action/keybinding tags resolve as expected. - Worker URL rewrite mock covered `/docs/`, `/docs.md`, `/docs/index.md`, extensionless docs routes, `.html` routes, `/docs/llms.txt`, and `/docs/sitemap.xml`. - High-effort adversarial subagent review found blockers around channel-aware redirects, shallow-checkout date fallback, file size, duplicated Markdown path inference, and process spawning. Those were addressed. Remaining Notes - Local `python -m http.server` does not emulate Cloudflare Pages pretty URLs, `_redirects`, or the docs-proxy Worker, so full local `afdocs` still cannot prove content negotiation end to end. - Existing production remains unchanged until this PR is deployed through the docs workflow. - Production baseline `npx afdocs check https://zed.dev/docs/ --fixes --verbose` still reports the original failures before this PR is deployed: 12 passed, 8 failed, 3 skipped. Release Notes: - Improved docs AI-readiness by adding machine-readable discovery, Markdown access, and freshness metadata. --------- Co-authored-by: Katie Geer <katie@zed.dev> Co-authored-by: Ben Kunkle <ben@zed.dev>
2026-07-08 20:54:51 +00:00
1 parent 4f6d05dde0
commit a103d8322b
@@ -0,0 +1,549 @@
+use anyhow::{Context, Result};
+use mdbook::BookItem;
+use mdbook::book::Book;
+use regex::Regex;
+use std::collections::HashMap;
+use std::path::{Path, PathBuf};
+use std::sync::OnceLock;
+
+use crate::FRONT_MATTER_COMMENT;
+
+#[derive(Debug)]
+pub(crate) struct DocsPage {
+    section: String,
+    title: String,
+    description: Option<String>,
+    pub(crate) source_path: PathBuf,
+    content: String,
+}
+
+pub(crate) fn write_ai_discovery_artifacts(
+    pages: &[DocsPage],
+    destination: &Path,
+    site_url: &str,
+) -> Result<()> {
+    copy_markdown_sources(destination, site_url, pages)?;
+    write_llms_txt(destination, site_url, pages)?;
+    write_sitemap_xml(destination, site_url, pages)?;
+    Ok(())
+}
+
+pub(crate) fn docs_pages(book: &Book) -> Result<Vec<DocsPage>> {
+    let mut pages = Vec::new();
+    let mut section = "Docs".to_string();
+    for item in book.iter() {
+        let BookItem::Chapter(chapter) = item else {
+            if let BookItem::PartTitle(part_title) = item {
+                section.clone_from(part_title);
+            }
+            continue;
+        };
+        let Some(source_path) = chapter.source_path.as_ref() else {
+            continue;
+        };
+        if source_path == Path::new("SUMMARY.md") {
+            continue;
+        }
+        pages.push(DocsPage {
+            section: section.clone(),
+            title: chapter.name.clone(),
+            description: docs_page_description(&chapter.content),
+            source_path: source_path.clone(),
+            content: chapter.content.clone(),
+        });
+    }
+    Ok(pages)
+}
+
+fn copy_markdown_sources(destination: &Path, site_url: &str, pages: &[DocsPage]) -> Result<()> {
+    for page in pages {
+        let destination = destination.join(&page.source_path);
+        if let Some(parent) = destination.parent() {
+            std::fs::create_dir_all(parent).with_context(|| {
+                format!("failed to create markdown destination {}", parent.display())
+            })?;
+        }
+        let contents = rewrite_docs_links(&markdown_source_contents(&page.content), site_url);
+        std::fs::write(
+            &destination,
+            add_llms_markdown_directive(&contents, site_url),
+        )
+        .with_context(|| {
+            format!(
+                "failed to write markdown page {} to {}",
+                page.source_path.display(),
+                destination.display()
+            )
+        })?;
+    }
+    let getting_started = destination.join("getting-started.md");
+    if getting_started.exists() {
+        std::fs::copy(&getting_started, destination.join("index.md"))
+            .context("failed to write index.md markdown alias")?;
+    }
+    Ok(())
+}
+
+fn markdown_source_contents(contents: &str) -> String {
+    front_matter_comment_regex()
+        .replace(contents, "")
+        .trim_start()
+        .to_string()
+}
+
+fn docs_page_description(contents: &str) -> Option<String> {
+    docs_page_metadata(contents).and_then(|metadata| {
+        metadata
+            .get("description")
+            .map(|description| {
+                description
+                    .trim()
+                    .trim_matches('"')
+                    .split_whitespace()
+                    .collect::<Vec<_>>()
+                    .join(" ")
+            })
+            .filter(|description| !description.is_empty())
+    })
+}
+
+fn docs_page_metadata(contents: &str) -> Option<HashMap<String, String>> {
+    let captures = front_matter_comment_regex().captures(contents)?;
+    serde_json::from_str(&captures[1]).ok()
+}
+
+fn front_matter_comment_regex() -> &'static Regex {
+    static FRONT_MATTER_COMMENT_REGEX: OnceLock<Regex> = OnceLock::new();
+    FRONT_MATTER_COMMENT_REGEX
+        .get_or_init(|| Regex::new(&FRONT_MATTER_COMMENT.replace("{}", "([^\\n]*)")).unwrap())
+}
+
+fn write_llms_txt(destination: &Path, site_url: &str, pages: &[DocsPage]) -> Result<()> {
+    let mut contents = String::new();
+    contents.push_str("# Zed Docs\n\n");
+    contents.push_str(
+        "> Official Zed documentation index with links to Markdown versions of each docs page.\n\n",
+    );
+    contents.push_str(
+        "Use these links for concise Markdown copies of Zed documentation pages. Each linked page mirrors the corresponding `/docs/*.html` page without site navigation or styling.\n\n",
+    );
+    let mut current_section = None;
+    for page in pages {
+        if current_section != Some(page.section.as_str()) {
+            if current_section.is_some() {
+                contents.push('\n');
+            }
+            contents.push_str("## ");
+            contents.push_str(&markdown_text(&page.section));
+            contents.push_str("\n\n");
+            current_section = Some(page.section.as_str());
+        }
+        contents.push_str("- [");
+        contents.push_str(&markdown_text(&page.title));
+        contents.push_str("](");
+        contents.push_str(&absolute_docs_url(site_url, &page.source_path));
+        contents.push(')');
+        if let Some(description) = &page.description {
+            contents.push_str(": ");
+            contents.push_str(&markdown_text(description));
+        }
+        contents.push('\n');
+    }
+    std::fs::write(destination.join("llms.txt"), contents).context("failed to write llms.txt")?;
+    Ok(())
+}
+
+fn markdown_text(text: &str) -> String {
+    text.replace('\\', "\\\\")
+        .replace('[', "\\[")
+        .replace(']', "\\]")
+}
+
+fn write_sitemap_xml(destination: &Path, site_url: &str, pages: &[DocsPage]) -> Result<()> {
+    let mut contents = String::new();
+    contents.push_str("<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n");
+    contents.push_str("<urlset xmlns=\"http://www.sitemaps.org/schemas/sitemap/0.9\">\n");
+    for page in pages {
+        contents.push_str("  <url><loc>");
+        contents.push_str(&xml_escape(&absolute_docs_url(
+            site_url,
+            &page.source_path.with_extension("html"),
+        )));
+        contents.push_str("</loc>");
+        contents.push_str("</url>\n");
+    }
+    contents.push_str("</urlset>\n");
+    std::fs::write(destination.join("sitemap.xml"), contents)
+        .context("failed to write sitemap.xml")?;
+    Ok(())
+}
+
+pub(crate) fn write_pages_redirects(
+    destination: &Path,
+    redirects: &[(String, String)],
+    site_url: &str,
+) -> Result<()> {
+    let Some(deploy_root) = destination.parent() else {
+        return Ok(());
+    };
+    let mut contents = String::new();
+    for (source, destination) in redirects {
+        write_redirect_line(
+            &mut contents,
+            &docs_path("/docs/", source),
+            &redirect_destination(site_url, destination),
+        );
+        if let Some(extensionless_source) = strip_html_suffix(source) {
+            write_redirect_line(
+                &mut contents,
+                &docs_path("/docs/", &extensionless_source),
+                &redirect_destination(
+                    site_url,
+                    &strip_html_suffix(destination).unwrap_or_else(|| destination.to_string()),
+                ),
+            );
+        }
+        if let Some(markdown_source) = html_path_to_markdown(source) {
+            if let Some(markdown_destination) = html_path_to_markdown(destination) {
+                write_redirect_line(
+                    &mut contents,
+                    &docs_path("/docs/", &markdown_source),
+                    &redirect_destination(site_url, &markdown_destination),
+                );
+            }
+        }
+    }
+    std::fs::write(deploy_root.join("_redirects"), contents)
+        .context("failed to write Cloudflare Pages _redirects")?;
+    Ok(())
+}
+
+pub(crate) fn write_markdown_redirect_aliases(
+    destination: &Path,
+    redirects: &[(String, String)],
+    site_url: &str,
+) -> Result<()> {
+    for (source, redirect_destination_path) in redirects {
+        let Some(source_markdown) = html_path_to_markdown(source) else {
+            continue;
+        };
+        let Some(destination_markdown) = html_path_to_markdown(redirect_destination_path) else {
+            continue;
+        };
+        let source_markdown = destination.join(source_markdown.trim_start_matches('/'));
+        let destination_markdown =
+            destination.join(destination_markdown.trim_start_matches("/docs/"));
+        if !destination_markdown.exists() {
+            continue;
+        }
+        if let Some(parent) = source_markdown.parent() {
+            std::fs::create_dir_all(parent).with_context(|| {
+                format!(
+                    "failed to create markdown alias directory {}",
+                    parent.display()
+                )
+            })?;
+        }
+        let contents = format!(
+            "# Moved\n\n> For the complete documentation index and Markdown links, see [llms.txt]({}).\n\nThis page moved to [the current docs page]({}).\n",
+            docs_url(site_url, Path::new("llms.txt")),
+            html_path_to_markdown(redirect_destination_path)
+                .map(|path| redirect_destination(site_url, &path))
+                .unwrap_or_else(|| redirect_destination(site_url, redirect_destination_path))
+        );
+        std::fs::write(&source_markdown, contents).with_context(|| {
+            format!(
+                "failed to write markdown redirect alias from {} to {}",
+                redirect_destination_path,
+                source_markdown.display()
+            )
+        })?;
+    }
+    Ok(())
+}
+
+fn write_redirect_line(contents: &mut String, source: &str, destination: &str) {
+    contents.push_str(source);
+    contents.push(' ');
+    contents.push_str(destination);
+    contents.push_str(" 301\n");
+}
+
+fn docs_path(site_url: &str, path: &str) -> String {
+    docs_url(site_url, Path::new(path.trim_start_matches('/')))
+}
+
+fn redirect_destination(site_url: &str, destination: &str) -> String {
+    if let Some(path) = destination.strip_prefix("/docs/") {
+        docs_url(site_url, Path::new(path))
+    } else if destination == "/docs" {
+        docs_url(site_url, Path::new(""))
+    } else {
+        destination.to_string()
+    }
+}
+
+fn strip_html_suffix(path: &str) -> Option<String> {
+    let (path, fragment) = split_fragment(path);
+    let path = path.strip_suffix(".html")?;
+    Some(format!("{path}{fragment}"))
+}
+
+fn html_path_to_markdown(path: &str) -> Option<String> {
+    let (path, fragment) = split_fragment(path);
+    if !path.starts_with("/docs/") && path != "/docs" && !path.ends_with(".html") {
+        return None;
+    }
+    let markdown_path = path.strip_suffix(".html").unwrap_or(path);
+    Some(format!("{markdown_path}.md{fragment}"))
+}
+
+fn split_fragment(path: &str) -> (&str, &str) {
+    match path.find('#') {
+        Some(index) => (&path[..index], &path[index..]),
+        None => (path, ""),
+    }
+}
+
+pub(crate) fn rewrite_docs_links(contents: &str, site_url: &str) -> String {
+    const STABLE_DOCS_PREFIX: &str = "https://zed.dev/docs/";
+    let channel_docs_prefix = absolute_docs_url(site_url, Path::new(""));
+    if channel_docs_prefix == STABLE_DOCS_PREFIX {
+        return contents.to_string();
+    }
+
+    let mut output = String::with_capacity(contents.len());
+    let mut remaining = contents;
+    while let Some(index) = remaining.find(STABLE_DOCS_PREFIX) {
+        output.push_str(&remaining[..index]);
+        let after_prefix = &remaining[index + STABLE_DOCS_PREFIX.len()..];
+        if after_prefix.starts_with("preview/") || after_prefix.starts_with("nightly/") {
+            output.push_str(STABLE_DOCS_PREFIX);
+        } else {
+            output.push_str(&channel_docs_prefix);
+        }
+        remaining = after_prefix;
+    }
+    output.push_str(remaining);
+    output
+}
+
+pub(crate) fn add_markdown_alternate_link(
+    contents: &str,
+    html_file: &Path,
+    root_dir: &Path,
+    site_url: &str,
+) -> String {
+    let Ok(relative_path) = html_file.strip_prefix(root_dir) else {
+        return contents.to_string();
+    };
+    let markdown_path = relative_path.with_extension("md");
+    if !root_dir.join(&markdown_path).exists() {
+        return contents.to_string();
+    }
+    let markdown_url = docs_url(site_url, &markdown_path);
+    let link = format!(
+        "        <link rel=\"alternate\" type=\"text/markdown\" href=\"{}\">\n",
+        markdown_url
+    );
+    contents.replacen("</head>", &(link + "    </head>"), 1)
+}
+
+fn add_llms_markdown_directive(contents: &str, site_url: &str) -> String {
+    let directive = format!(
+        "> For the complete documentation index and Markdown links, see [llms.txt]({}).\n\n",
+        docs_url(site_url, Path::new("llms.txt")),
+    );
+    if let Some(rest) = contents.strip_prefix("---\n") {
+        if let Some(frontmatter_end) = rest.find("\n---\n") {
+            let split_at = "---\n".len() + frontmatter_end + "\n---\n".len();
+            let mut output = String::with_capacity(contents.len() + directive.len());
+            output.push_str(&contents[..split_at]);
+            output.push('\n');
+            output.push_str(&directive);
+            output.push_str(&contents[split_at..]);
+            return output;
+        }
+    }
+
+    let mut output = String::with_capacity(contents.len() + directive.len());
+    output.push_str(&directive);
+    output.push_str(contents);
+    output
+}
+
+fn docs_url(site_url: &str, path: &Path) -> String {
+    let mut url = site_url.to_string();
+    if !url.ends_with('/') {
+        url.push('/');
+    }
+    url.push_str(&path.to_string_lossy().replace('\\', "/"));
+    url
+}
+
+fn absolute_docs_url(site_url: &str, path: &Path) -> String {
+    let url = docs_url(site_url, path);
+    if url.starts_with("http://") || url.starts_with("https://") {
+        url
+    } else {
+        format!("https://zed.dev{}", url)
+    }
+}
+
+fn xml_escape(value: &str) -> String {
+    value
+        .replace('&', "&amp;")
+        .replace('<', "&lt;")
+        .replace('>', "&gt;")
+        .replace('"', "&quot;")
+        .replace('\'', "&apos;")
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_add_llms_markdown_directive_inserts_after_frontmatter() {
+        let contents = "---\ntitle: Example\n---\n# Example\n";
+        let output = add_llms_markdown_directive(contents, "/docs/");
+
+        assert!(output.starts_with("---\ntitle: Example\n---\n\n"));
+        assert!(output.contains(
+            "> For the complete documentation index and Markdown links, see [llms.txt](/docs/llms.txt)."
+        ));
+    }
+
+    #[test]
+    fn test_redirect_destination_uses_channel_site_url_for_docs_paths() {
+        assert_eq!(
+            redirect_destination("/docs/preview/", "/docs/ai/overview.html"),
+            "/docs/preview/ai/overview.html"
+        );
+        assert_eq!(
+            redirect_destination("/docs/preview/", "/community-links"),
+            "/community-links"
+        );
+    }
+
+    #[test]
+    fn test_rewrite_docs_links_uses_channel_site_url() {
+        assert_eq!(
+            rewrite_docs_links(
+                "See [Code Actions](https://zed.dev/docs/configuring-languages#code-actions) and [Preview](https://zed.dev/docs/preview/ai/overview.html).",
+                "/docs/preview/"
+            ),
+            "See [Code Actions](https://zed.dev/docs/preview/configuring-languages#code-actions) and [Preview](https://zed.dev/docs/preview/ai/overview.html)."
+        );
+    }
+
+    #[test]
+    fn test_docs_path_uses_channel_site_url() {
+        assert_eq!(
+            docs_path("/docs/preview/", "/assistant.md"),
+            "/docs/preview/assistant.md"
+        );
+    }
+
+    #[test]
+    fn test_write_pages_redirects_keeps_sources_on_internal_pages_path() -> Result<()> {
+        let deploy_root = std::env::temp_dir().join(format!(
+            "docs_preprocessor_pages_redirects_test_{}_{}",
+            std::process::id(),
+            std::time::SystemTime::now()
+                .duration_since(std::time::UNIX_EPOCH)?
+                .as_nanos()
+        ));
+        let destination = deploy_root.join("docs");
+        std::fs::create_dir_all(&destination)?;
+        let redirects = vec![
+            (
+                "/assistant.html".to_string(),
+                "/docs/ai/overview.html".to_string(),
+            ),
+            (
+                "/community/feedback.html".to_string(),
+                "/community-links".to_string(),
+            ),
+        ];
+
+        write_pages_redirects(&destination, &redirects, "/docs/preview/")?;
+
+        assert_eq!(
+            std::fs::read_to_string(deploy_root.join("_redirects"))?,
+            "/docs/assistant.html /docs/preview/ai/overview.html 301\n\
+/docs/assistant /docs/preview/ai/overview 301\n\
+/docs/assistant.md /docs/preview/ai/overview.md 301\n\
+/docs/community/feedback.html /community-links 301\n\
+/docs/community/feedback /community-links 301\n"
+        );
+        std::fs::remove_dir_all(&deploy_root)?;
+        Ok(())
+    }
+
+    #[test]
+    fn test_write_ai_discovery_artifacts_generates_agent_facing_metadata() -> Result<()> {
+        let destination = std::env::temp_dir().join(format!(
+            "docs_preprocessor_ai_discovery_test_{}_{}",
+            std::process::id(),
+            std::time::SystemTime::now()
+                .duration_since(std::time::UNIX_EPOCH)?
+                .as_nanos()
+        ));
+        std::fs::create_dir_all(&destination)?;
+
+        let pages = vec![
+            DocsPage {
+                section: "Docs".to_string(),
+                title: "Getting Started".to_string(),
+                description: Some("Start using Zed.".to_string()),
+                source_path: PathBuf::from("getting-started.md"),
+                content: format!(
+                    "{}\n# Getting Started\n",
+                    FRONT_MATTER_COMMENT.replace("{}", r#"{"description":"Start using Zed."}"#)
+                ),
+            },
+            DocsPage {
+                section: "AI".to_string(),
+                title: "MCP".to_string(),
+                description: Some("Connect model context servers.".to_string()),
+                source_path: PathBuf::from("ai/mcp.md"),
+                content: format!(
+                    "{}\n# MCP\n",
+                    FRONT_MATTER_COMMENT
+                        .replace("{}", r#"{"description":"Connect model context servers."}"#)
+                ),
+            },
+        ];
+
+        write_ai_discovery_artifacts(&pages, &destination, "/docs/")?;
+
+        let llms_txt = std::fs::read_to_string(destination.join("llms.txt"))?;
+        assert!(llms_txt.contains("## Docs"));
+        assert!(llms_txt.contains(
+            "- [Getting Started](https://zed.dev/docs/getting-started.md): Start using Zed."
+        ));
+        assert!(llms_txt.contains("## AI"));
+        assert!(
+            llms_txt.contains(
+                "- [MCP](https://zed.dev/docs/ai/mcp.md): Connect model context servers."
+            )
+        );
+
+        let sitemap_xml = std::fs::read_to_string(destination.join("sitemap.xml"))?;
+        assert!(sitemap_xml.contains("<loc>https://zed.dev/docs/getting-started.html</loc>"));
+        assert!(sitemap_xml.contains("<loc>https://zed.dev/docs/ai/mcp.html</loc>"));
+
+        let mcp_markdown = std::fs::read_to_string(destination.join("ai/mcp.md"))?;
+        assert!(mcp_markdown.starts_with(
+            "> For the complete documentation index and Markdown links, see [llms.txt](/docs/llms.txt).\n\n# MCP"
+        ));
+        assert!(!mcp_markdown.contains("ZED_META"));
+
+        let index_markdown = std::fs::read_to_string(destination.join("index.md"))?;
+        assert!(index_markdown.contains("# Getting Started"));
+
+        std::fs::remove_dir_all(&destination)?;
+        Ok(())
+    }
+}
@@ -1,4 +1,10 @@
 use anyhow::{Context, Result};
+mod ai_discovery;
+
+use ai_discovery::{
+    add_markdown_alternate_link, docs_pages, rewrite_docs_links, write_ai_discovery_artifacts,
+    write_markdown_redirect_aliases, write_pages_redirects,
+};
 use mdbook::BookItem;
 use mdbook::book::{Book, Chapter};
 use mdbook::preprocess::CmdPreprocessor;
@@ -188,7 +194,7 @@ fn handle_preprocessing() -> Result<()> {
    template_big_table_of_actions(&mut book);
    template_and_validate_keybindings(&mut book, &mut errors);
    template_and_validate_actions(&mut book, &mut errors);
-    template_and_validate_json_snippets(&mut book, &mut errors);
+    template_and_validate_json_snippets(&mut book, &mut errors)?;

    if !errors.is_empty() {
        const ANSI_RED: &str = "\x1b[31m";
@@ -252,7 +258,7 @@ fn format_binding(binding: String) -> String {
 }

 fn template_and_validate_keybindings(book: &mut Book, errors: &mut HashSet<PreprocessorError>) {
-    let regex = Regex::new(r"\{#kb(?::(\w+))?\s+(.*?)\}").unwrap();
+    let regex = Regex::new(r"(?s)\{#kb(?::(\w+))?\s+(.*?)\}").unwrap();

    for_each_chapter_mut(book, |chapter| {
        chapter.content = regex
@@ -300,7 +306,7 @@ fn template_and_validate_keybindings(book: &mut Book, errors: &mut HashSet<Prepr
 }

 fn template_and_validate_actions(book: &mut Book, errors: &mut HashSet<PreprocessorError>) {
-    let regex = Regex::new(r"\{#action (.*?)\}").unwrap();
+    let regex = Regex::new(r"(?s)\{#action\s+(.*?)\}").unwrap();

    for_each_chapter_mut(book, |chapter| {
        chapter.content = regex
@@ -379,7 +385,10 @@ fn find_binding_with_overlay(
        .or_else(|| find_binding(os, action))
 }

-fn template_and_validate_json_snippets(book: &mut Book, errors: &mut HashSet<PreprocessorError>) {
+fn template_and_validate_json_snippets(
+    book: &mut Book,
+    errors: &mut HashSet<PreprocessorError>,
+) -> Result<()> {
    let params = SettingsJsonSchemaParams {
        language_names: &[],
        font_names: &[],
@@ -393,7 +402,7 @@ fn template_and_validate_json_snippets(book: &mut Book, errors: &mut HashSet<Pre
    };
    let settings_schema = SettingsStore::json_schema(&params);
    let settings_validator = jsonschema::validator_for(&settings_schema)
-        .expect("failed to compile settings JSON schema");
+        .context("failed to compile settings JSON schema")?;

    // The keymap schema is built from the action manifest. When `actions.json`
    // is unavailable (e.g. when running outside of CI without first generating
@@ -405,7 +414,7 @@ fn template_and_validate_json_snippets(book: &mut Book, errors: &mut HashSet<Pre
            keymap_schema_for_actions(&ALL_ACTIONS.actions, &ALL_ACTIONS.schema_definitions);
        Some(
            jsonschema::validator_for(&keymap_schema)
-                .expect("failed to compile keymap JSON schema"),
+                .context("failed to compile keymap JSON schema")?,
        )
    } else {
        None
@@ -568,6 +577,8 @@ fn template_and_validate_json_snippets(book: &mut Book, errors: &mut HashSet<Pre
        };
        Ok(())
    });
+
+    Ok(())
 }

 /// Removes any configurable options from the stringified action if existing,
@@ -679,6 +690,19 @@ fn handle_postprocessing() -> Result<()> {
        .as_table_mut()
        .expect("output is table");
    let zed_html = output.remove("zed-html").expect("zed-html output defined");
+    let redirects = zed_html
+        .get("redirect")
+        .and_then(|redirects| redirects.as_table())
+        .map(|redirects| {
+            redirects
+                .iter()
+                .filter_map(|(source, destination)| {
+                    destination
+                        .as_str()
+                        .map(|destination| (source.clone(), destination.to_string()))
+                })
+                .collect::<Vec<_>>()
+        });
    let default_description = zed_html
        .get("default-description")
        .expect("Default description not found")
@@ -694,6 +718,17 @@ fn handle_postprocessing() -> Result<()> {
    let amplitude_key = std::env::var("DOCS_AMPLITUDE_API_KEY").unwrap_or_default();
    let consent_io_instance = std::env::var("DOCS_CONSENT_IO_INSTANCE").unwrap_or_default();
    let docs_channel = std::env::var("DOCS_CHANNEL").unwrap_or_else(|_| "stable".to_string());
+    let site_url = std::env::var("MDBOOK_BOOK__SITE_URL")
+        .ok()
+        .filter(|site_url| !site_url.trim().is_empty())
+        .unwrap_or_else(|| {
+            match docs_channel.as_str() {
+                "nightly" => "/docs/nightly/",
+                "preview" => "/docs/preview/",
+                _ => "/docs/",
+            }
+            .to_string()
+        });
    let noindex = if docs_channel == "nightly" || docs_channel == "preview" {
        "<meta name=\"robots\" content=\"noindex, nofollow\">"
    } else {
@@ -733,8 +768,10 @@ fn handle_postprocessing() -> Result<()> {
    }

    zlog::info!(logger => "Processing {} `.html` files", files.len());
+    let pages = docs_pages(&ctx.book)?;
+    write_ai_discovery_artifacts(&pages, &root_dir, &site_url)?;
    let meta_regex = Regex::new(&FRONT_MATTER_COMMENT.replace("{}", "(.*)")).unwrap();
-    for file in files {
+    for file in &files {
        let contents = std::fs::read_to_string(&file)?;
        let mut meta_description = None;
        let mut meta_title = None;
@@ -770,14 +807,19 @@ fn handle_postprocessing() -> Result<()> {
        let contents = contents.replace("#amplitude_key#", &amplitude_key);
        let contents = contents.replace("#consent_io_instance#", &consent_io_instance);
        let contents = contents.replace("#noindex#", noindex);
+        let contents = rewrite_docs_links(&contents, &site_url);
+        let contents = add_markdown_alternate_link(&contents, file, &root_dir, &site_url);
        let contents = title_regex()
            .replace(&contents, |_: &regex::Captures| {
                format!("<title>{}</title>", meta_title)
            })
            .to_string();
-        // let contents = contents.replace("#title#", &meta_title);
        std::fs::write(file, contents)?;
    }
+    if let Some(redirects) = redirects {
+        write_markdown_redirect_aliases(&root_dir, &redirects, &site_url)?;
+        write_pages_redirects(&root_dir, &redirects, &site_url)?;
+    }
    return Ok(());

    fn pretty_path<'a>(
@@ -906,66 +948,4 @@ fn keymap_schema_for_actions(
 }

 #[cfg(test)]
-mod tests {
-    use super::*;
-    use serde_json::json;
-
-    #[test]
-    fn test_find_binding_prefers_exact_match_over_parameterized() {
-        let keymap: KeymapFile = serde_json::from_value(json!([
-            {
-                "bindings": {
-                    "ctrl-tab": "agents_sidebar::ToggleThreadSwitcher",
-                    "ctrl-shift-tab": ["agents_sidebar::ToggleThreadSwitcher", { "select_last": true }]
-                }
-            }
-        ]))
-        .unwrap();
-
-        let binding = find_binding_in_keymap(&keymap, "agents_sidebar::ToggleThreadSwitcher");
-        assert_eq!(binding.as_deref(), Some("ctrl-tab"));
-    }
-
-    #[test]
-    fn test_find_binding_falls_back_to_parameterized_match() {
-        let keymap: KeymapFile = serde_json::from_value(json!([
-            {
-                "bindings": {
-                    "ctrl-shift-tab": ["agents_sidebar::ToggleThreadSwitcher", { "select_last": true }]
-                }
-            }
-        ]))
-        .unwrap();
-
-        let binding = find_binding_in_keymap(&keymap, "agents_sidebar::ToggleThreadSwitcher");
-        assert_eq!(binding.as_deref(), Some("ctrl-shift-tab"));
-    }
-
-    #[test]
-    fn test_find_binding_prefers_exact_match_regardless_of_order() {
-        let keymap: KeymapFile = serde_json::from_value(json!([
-            {
-                "bindings": {
-                    "ctrl-shift-tab": ["agents_sidebar::ToggleThreadSwitcher", { "select_last": true }],
-                    "ctrl-tab": "agents_sidebar::ToggleThreadSwitcher"
-                }
-            }
-        ]))
-        .unwrap();
-
-        let binding = find_binding_in_keymap(&keymap, "agents_sidebar::ToggleThreadSwitcher");
-        assert_eq!(binding.as_deref(), Some("ctrl-tab"));
-    }
-
-    #[test]
-    fn test_find_binding_later_section_overrides_earlier() {
-        let keymap: KeymapFile = serde_json::from_value(json!([
-            { "bindings": { "ctrl-a": "some::Action" } },
-            { "bindings": { "ctrl-b": "some::Action" } }
-        ]))
-        .unwrap();
-
-        let binding = find_binding_in_keymap(&keymap, "some::Action");
-        assert_eq!(binding.as_deref(), Some("ctrl-b"));
-    }
-}
+mod tests;
@@ -0,0 +1,61 @@
+use super::*;
+use serde_json::json;
+
+#[test]
+fn test_find_binding_prefers_exact_match_over_parameterized() {
+    let keymap: KeymapFile = serde_json::from_value(json!([
+        {
+            "bindings": {
+                "ctrl-tab": "agents_sidebar::ToggleThreadSwitcher",
+                "ctrl-shift-tab": ["agents_sidebar::ToggleThreadSwitcher", { "select_last": true }]
+            }
+        }
+    ]))
+    .unwrap();
+
+    let binding = find_binding_in_keymap(&keymap, "agents_sidebar::ToggleThreadSwitcher");
+    assert_eq!(binding.as_deref(), Some("ctrl-tab"));
+}
+
+#[test]
+fn test_find_binding_falls_back_to_parameterized_match() {
+    let keymap: KeymapFile = serde_json::from_value(json!([
+        {
+            "bindings": {
+                "ctrl-shift-tab": ["agents_sidebar::ToggleThreadSwitcher", { "select_last": true }]
+            }
+        }
+    ]))
+    .unwrap();
+
+    let binding = find_binding_in_keymap(&keymap, "agents_sidebar::ToggleThreadSwitcher");
+    assert_eq!(binding.as_deref(), Some("ctrl-shift-tab"));
+}
+
+#[test]
+fn test_find_binding_prefers_exact_match_regardless_of_order() {
+    let keymap: KeymapFile = serde_json::from_value(json!([
+        {
+            "bindings": {
+                "ctrl-shift-tab": ["agents_sidebar::ToggleThreadSwitcher", { "select_last": true }],
+                "ctrl-tab": "agents_sidebar::ToggleThreadSwitcher"
+            }
+        }
+    ]))
+    .unwrap();
+
+    let binding = find_binding_in_keymap(&keymap, "agents_sidebar::ToggleThreadSwitcher");
+    assert_eq!(binding.as_deref(), Some("ctrl-tab"));
+}
+
+#[test]
+fn test_find_binding_later_section_overrides_earlier() {
+    let keymap: KeymapFile = serde_json::from_value(json!([
+        { "bindings": { "ctrl-a": "some::Action" } },
+        { "bindings": { "ctrl-b": "some::Action" } }
+    ]))
+    .unwrap();
+
+    let binding = find_binding_in_keymap(&keymap, "some::Action");
+    assert_eq!(binding.as_deref(), Some("ctrl-b"));
+}