feat: uBlock Origin Lite integration for ad-blocking during WebPage captures

- singlefile.rs: when ARCHIVR_UBLOCK=true and ARCHIVR_UBLOCK_EXT is set,
  archivr owns Chrome's lifecycle (--headless=new, --remote-debugging-port,
  --load-extension); single-file connects via --browser-server instead of
  launching its own Chrome. Falls back to old behaviour with ublock_skipped=true
  when the ext path is missing or invalid.
- capture.rs: thread ublock_skipped through CaptureResult
- database.rs: add notes_json TEXT column to capture_jobs (DDL + idempotent
  ALTER TABLE migration); update_capture_job_status gains notes_json param
- archive.rs: expose notes_json in CaptureJobSummary
- routes.rs: store {"ublock_skipped":true} in notes_json on completed captures
- ToastStack.jsx: warning toast variant (toast--warning) with Details expander
  and Ignore button
- CaptureDialog.jsx: fire warning toast when poll result has ublock_skipped
- App.jsx: sessionStorage-backed Ignore suppression for ublock warnings
- styles.css: .toast--warning (amber left border) + .toast-warning-detail
- flake.nix: ublockLite derivation fetches uBOLite_2026.705.2152.chromium.zip
  (pinned SHA256) from uBlockOrigin/uBOL-home; sets ARCHIVR_UBLOCK_EXT in both
  archivr and archivr-server wrappers

Env vars:
  ARCHIVR_UBLOCK=true (default) — enable uBlock during WebPage captures
  ARCHIVR_UBLOCK_EXT — path to unpacked uBOL extension dir (set by Nix)

crates/archivr-core/src/archive.rs

@@ -69,6 +69,7 @@ pub struct CaptureJobSummary {
     pub run_uid: Option<String>,
     pub status: String,
     pub error_text: Option<String>,
+    pub notes_json: Option<String>,
     pub created_at: String,
     pub updated_at: String,
 }
@@ -341,6 +342,7 @@ pub fn get_capture_job(
         run_uid: r.run_uid,
         status: r.status,
         error_text: r.error_text,
+        notes_json: r.notes_json,
         created_at: r.created_at,
         updated_at: r.updated_at,
     }))

crates/archivr-core/src/capture.rs

@@ -37,6 +37,9 @@ pub enum Source {
 pub struct CaptureResult {
     pub run_uid: String,
     pub status: String,
+    /// `true` when uBlock was requested but the extension path was not found.
+    /// The capture succeeded without ad-blocking; the UI should warn the user.
+    pub ublock_skipped: bool,
 }
 
 #[derive(Debug, Clone, Default)]
@@ -991,6 +994,7 @@ pub fn perform_capture(
                 return Ok(CaptureResult {
                     run_uid: run.run_uid.clone(),
                     status: "completed".to_string(),
+                    ublock_skipped: false,
                 });
             }
             Err(e) => {
@@ -1133,6 +1137,7 @@ pub fn perform_capture(
                 return Ok(CaptureResult {
                     run_uid: run.run_uid.clone(),
                     status: "completed".to_string(),
+                    ublock_skipped: result.ublock_skipped,
                 });
             }
             Err(e) => {
@@ -1187,6 +1192,7 @@ pub fn perform_capture(
                 return Ok(CaptureResult {
                     run_uid: run.run_uid.clone(),
                     status: "completed".to_string(),
+                    ublock_skipped: false,
                 });
             }
             Err(e) => {
@@ -1342,6 +1348,7 @@ pub fn perform_capture(
     Ok(CaptureResult {
         run_uid: run.run_uid.clone(),
         status: "completed".to_string(),
+        ublock_skipped: false,
     })
 }
 

crates/archivr-core/src/database.rs

@@ -105,6 +105,7 @@ pub struct CaptureJobRecord {
     pub run_uid: Option<String>,
     pub status: String,
     pub error_text: Option<String>,
+    pub notes_json: Option<String>,
     pub created_at: String,
     pub updated_at: String,
 }
@@ -308,6 +309,7 @@ pub fn initialize_schema(conn: &Connection) -> Result<()> {
             run_uid     TEXT,
             status      TEXT NOT NULL CHECK (status IN ('pending','running','completed','failed')) DEFAULT 'pending',
             error_text  TEXT,
+            notes_json  TEXT,
             created_at  TEXT NOT NULL,
             updated_at  TEXT NOT NULL
         );
@@ -393,6 +395,10 @@ pub fn initialize_schema(conn: &Connection) -> Result<()> {
         )?;
     }
 
+    // Migration: add notes_json column to existing capture_jobs tables.
+    // Silently ignored when the column already exists (idempotent).
+    let _ = conn.execute("ALTER TABLE capture_jobs ADD COLUMN notes_json TEXT", []);
+
     Ok(())
 }
 
@@ -1142,19 +1148,21 @@ pub fn create_capture_job(conn: &Connection, archive_id: &str) -> Result<String>
     Ok(job_uid)
 }
 
-/// Updates the status (and optionally run_uid / error_text) of a capture job.
+/// Updates the status (and optionally run_uid / error_text / notes_json) of a capture job.
 pub fn update_capture_job_status(
     conn: &Connection,
     job_uid: &str,
     status: &str,
     run_uid: Option<&str>,
     error_text: Option<&str>,
+    notes_json: Option<&str>,
 ) -> Result<()> {
     let now = now_timestamp();
     conn.execute(
         "UPDATE capture_jobs SET status = ?1, run_uid = COALESCE(?2, run_uid),
-         error_text = ?3, updated_at = ?4 WHERE job_uid = ?5",
-        rusqlite::params![status, run_uid, error_text, now, job_uid],
+         error_text = ?3, notes_json = COALESCE(?4, notes_json), updated_at = ?5
+         WHERE job_uid = ?6",
+        rusqlite::params![status, run_uid, error_text, notes_json, now, job_uid],
     )?;
     Ok(())
 }
@@ -1162,7 +1170,7 @@ pub fn update_capture_job_status(
 /// Returns a capture job by uid.
 pub fn get_capture_job(conn: &Connection, job_uid: &str) -> Result<Option<CaptureJobRecord>> {
     conn.query_row(
-        "SELECT job_uid, archive_id, run_uid, status, error_text, created_at, updated_at
+        "SELECT job_uid, archive_id, run_uid, status, error_text, notes_json, created_at, updated_at
          FROM capture_jobs WHERE job_uid = ?1",
         [job_uid],
         |row| {
@@ -1172,8 +1180,9 @@ pub fn get_capture_job(conn: &Connection, job_uid: &str) -> Result<Option<Captur
                 run_uid: row.get(2)?,
                 status: row.get(3)?,
                 error_text: row.get(4)?,
-                created_at: row.get(5)?,
-                updated_at: row.get(6)?,
+                notes_json: row.get(5)?,
+                created_at: row.get(6)?,
+                updated_at: row.get(7)?,
             })
         },
     )
@@ -2806,8 +2815,8 @@ mod tests {
     fn capture_job_status_transitions() {
         let conn = conn();
         let job_uid = create_capture_job(&conn, "test").unwrap();
-        update_capture_job_status(&conn, &job_uid, "running", None, None).unwrap();
-        update_capture_job_status(&conn, &job_uid, "completed", Some("run_abc"), None).unwrap();
+        update_capture_job_status(&conn, &job_uid, "running", None, None, None).unwrap();
+        update_capture_job_status(&conn, &job_uid, "completed", Some("run_abc"), None, None).unwrap();
         let job = get_capture_job(&conn, &job_uid).unwrap().unwrap();
         assert_eq!(job.status, "completed");
         assert_eq!(job.run_uid.as_deref(), Some("run_abc"));
@@ -2819,7 +2828,7 @@ mod tests {
 
         // Simulate an in-progress capture_job (run_uid still NULL — common crash case).
         let uid = create_capture_job(&conn, "test").unwrap();
-        update_capture_job_status(&conn, &uid, "running", None, None).unwrap();
+        update_capture_job_status(&conn, &uid, "running", None, None, None).unwrap();
 
         // Simulate an in-progress archive_run and item with no associated capture_job
         // (covers the case where run_uid was never written back before the crash).
@@ -3177,7 +3186,7 @@ mod tests {
     fn has_active_capture_jobs_true_for_running() {
         let conn = conn();
         let uid = create_capture_job(&conn, "test").unwrap();
-        update_capture_job_status(&conn, &uid, "running", None, None).unwrap();
+        update_capture_job_status(&conn, &uid, "running", None, None, None).unwrap();
         assert!(has_active_capture_jobs(&conn).unwrap());
     }
 
@@ -3185,7 +3194,7 @@ mod tests {
     fn has_active_capture_jobs_false_for_completed() {
         let conn = conn();
         let uid = create_capture_job(&conn, "test").unwrap();
-        update_capture_job_status(&conn, &uid, "completed", Some("run_x"), None).unwrap();
+        update_capture_job_status(&conn, &uid, "completed", Some("run_x"), None, None).unwrap();
         assert!(!has_active_capture_jobs(&conn).unwrap());
     }
 

crates/archivr-core/src/downloader/singlefile.rs

@@ -1,7 +1,15 @@
 use anyhow::{Context, Result, bail};
 use base64::engine::general_purpose::STANDARD as B64;
 use base64::Engine as _;
-use std::{collections::HashMap, env, io::Read, path::Path, process::Command};
+use std::{
+    collections::HashMap,
+    env,
+    io::Read,
+    net::TcpListener,
+    path::{Path, PathBuf},
+    process::Command,
+    time::{Duration, Instant},
+};
 
 use crate::downloader::cookies::{domain_from_url, write_netscape_cookie_file};
 use crate::hash::hash_file;
@@ -17,26 +25,88 @@ pub struct SaveResult {
     pub favicon_hash: Option<String>,
     /// File extension for the favicon (e.g. `".ico"`, `".png"`), if present.
     pub favicon_ext: Option<String>,
+    /// `true` when `ARCHIVR_UBLOCK=true` (the default) but the extension path
+    /// was missing or invalid.  The capture succeeded but ran without ad-blocking.
+    pub ublock_skipped: bool,
 }
 
 /// Archives `url` as a self-contained HTML snapshot.
 ///
-/// Returns `(sha256_hex, title_hint)` on success.
-/// - `sha256_hex`: hash of the saved `.html` file, used as the blob key.
-/// - `title_hint`: page title extracted from the `<title>` tag, if present.
-///
-/// Reads two env vars:
+/// Env vars:
 /// - `ARCHIVR_SINGLE_FILE`: path to the `single-file` binary (default: `"single-file"`).
 /// - `ARCHIVR_CHROME`: path to the Chromium/Chrome binary (default: `"chromium"`).
-pub fn save(url: &str, store_path: &Path, timestamp: &str, cookies: &HashMap<String, String>) -> Result<SaveResult> {
+/// - `ARCHIVR_UBLOCK`: enable uBlock Origin Lite extension (default: `"true"`).
+/// - `ARCHIVR_UBLOCK_EXT`: path to the unpacked uBlock Origin Lite extension directory.
+/// - `ARCHIVR_CHROME_ARGS`: space-separated extra Chrome flags (e.g. `"--no-sandbox"`).
+pub fn save(
+    url: &str,
+    store_path: &Path,
+    timestamp: &str,
+    cookies: &HashMap<String, String>,
+) -> Result<SaveResult> {
     let single_file =
         env::var("ARCHIVR_SINGLE_FILE").unwrap_or_else(|_| "single-file".to_string());
     let chrome = env::var("ARCHIVR_CHROME").unwrap_or_else(|_| "chromium".to_string());
-    save_with(url, store_path, timestamp, &single_file, &chrome, cookies)
+    let (ublock_ext, ublock_skipped) = resolve_ublock_config();
+    let mut result = save_with(
+        url,
+        store_path,
+        timestamp,
+        &single_file,
+        &chrome,
+        cookies,
+        ublock_ext.as_deref(),
+    )?;
+    result.ublock_skipped = ublock_skipped;
+    Ok(result)
 }
 
-/// Inner implementation; takes binary paths explicitly so tests can inject them
-/// without mutating process-global environment variables.
+/// Reads `ARCHIVR_UBLOCK` and `ARCHIVR_UBLOCK_EXT` and returns:
+///
+/// - `(Some(path), false)` — uBlock is enabled and the extension dir is valid.
+/// - `(None, true)`  — uBlock is enabled but the extension dir is missing/invalid
+///                     (warns to stderr; caller should surface this to the user).
+/// - `(None, false)` — uBlock is explicitly disabled (`ARCHIVR_UBLOCK=false/0`).
+fn resolve_ublock_config() -> (Option<PathBuf>, bool) {
+    let enabled = env::var("ARCHIVR_UBLOCK").unwrap_or_else(|_| "true".to_string());
+    if enabled.eq_ignore_ascii_case("false") || enabled == "0" {
+        return (None, false);
+    }
+    match env::var("ARCHIVR_UBLOCK_EXT").ok().filter(|s| !s.is_empty()) {
+        None => {
+            eprintln!(
+                "warn: uBlock: ARCHIVR_UBLOCK=true but ARCHIVR_UBLOCK_EXT is not set; \
+                 capturing without ad-blocking"
+            );
+            (None, true)
+        }
+        Some(ext_path_str) => {
+            let path = PathBuf::from(&ext_path_str);
+            if path.is_dir() {
+                (Some(path), false)
+            } else {
+                eprintln!(
+                    "warn: uBlock: ARCHIVR_UBLOCK_EXT={ext_path_str:?} is not a directory; \
+                     capturing without ad-blocking"
+                );
+                (None, true)
+            }
+        }
+    }
+}
+
+/// Inner implementation.  Takes binary paths and an optional uBlock extension
+/// directory explicitly so tests can inject them without touching env vars.
+///
+/// When `ublock_ext` is `Some(path)` we own Chrome's lifecycle:
+///   1. allocate a free TCP port,
+///   2. launch Chrome headless with the extension loaded,
+///   3. wait for Chrome's DevTools HTTP API to respond,
+///   4. run single-file pointing at our Chrome via `--browser-server`,
+///   5. kill Chrome after single-file exits.
+///
+/// When `ublock_ext` is `None` the original behaviour is preserved:
+/// single-file launches and manages Chrome internally.
 fn save_with(
     url: &str,
     store_path: &Path,
@@ -44,17 +114,18 @@ fn save_with(
     single_file: &str,
     chrome: &str,
     cookies: &HashMap<String, String>,
+    ublock_ext: Option<&Path>,
 ) -> Result<SaveResult> {
     let temp_dir = store_path.join("temp").join(timestamp);
     std::fs::create_dir_all(&temp_dir).context("failed to create temp dir")?;
 
     let out_file = temp_dir.join(format!("{timestamp}.html"));
 
-    // Write a user script that strips <script> elements from the live DOM
-    // just before SingleFile serializes it. This lets scripts execute during
-    // capture (so JS-applied CSS is present) without leaving data:-URL ES
-    // modules in the saved file that would cause "base scheme isn't
-    // hierarchical" errors in the viewer. JSON-LD structured data is kept.
+    // User script that strips <script> elements from the live DOM just before
+    // SingleFile serializes it.  Lets scripts execute during capture (so
+    // JS-applied CSS is present) without leaving data:-URL ES modules in the
+    // saved file that would cause "base scheme isn't hierarchical" errors.
+    // JSON-LD structured data is kept.
     let user_script_path = temp_dir.join("sf-strip-scripts.js");
     std::fs::write(
         &user_script_path,
@@ -65,38 +136,21 @@ fn save_with(
     )
     .context("failed to write single-file user script")?;
 
-    // Chrome's user-data-dir for this capture. Required alongside
-    // --disable-web-security — newer Chromium silently ignores that flag
-    // without a writable user-data-dir. Using a subdirectory of temp_dir
-    // keeps it isolated and it gets cleaned up with the rest of the temp dir.
+    // Isolated Chrome profile directory; keeps each capture's browser state
+    // separate and cleans up with the rest of the temp dir.
     let chrome_data_dir = temp_dir.join("chrome-data");
-    // Build the browser-args JSON array. Start with the flags always required,
-    // then append any extra flags from ARCHIVR_CHROME_ARGS (space-separated).
-    // Docker containers running as root need "--no-sandbox" here because
-    // Chromium refuses to start as root without it.
-    //
-    // --window-size is set to a realistic desktop viewport so that
-    // --remove-alternative-medias=false and --remove-unused-styles=false
-    // actually preserve responsive @media rules and styles that only match
-    // at normal screen widths (headless Chromium defaults to a small viewport
-    // that would otherwise defeat the preservation flags).
-    let mut chrome_flags = vec![
-        "--disable-web-security".to_string(),
-        format!("--user-data-dir={}", chrome_data_dir.display()),
-        "--window-size=1920,1080".to_string(),
-    ];
-    if let Ok(extra) = std::env::var("ARCHIVR_CHROME_ARGS") {
-        chrome_flags.extend(extra.split_whitespace().filter(|s| !s.is_empty()).map(str::to_string));
-    }
-    let quoted: Vec<String> = chrome_flags
-        .iter()
-        .map(|f| format!("\"{}\"", f.replace('\\', "\\\\").replace('"', "\\\"")))
-        .collect();
-    let browser_args = format!("[{}]", quoted.join(","));
 
-    // Write cookie file if cookies are provided.
+    // Extra Chrome flags from the environment (e.g. "--no-sandbox" in Docker).
+    let extra_chrome_args: Vec<String> = env::var("ARCHIVR_CHROME_ARGS")
+        .unwrap_or_default()
+        .split_whitespace()
+        .filter(|s| !s.is_empty())
+        .map(str::to_string)
+        .collect();
+
+    // Write cookie file before running Chrome so it's available immediately.
     // Never pass cookie values in process args (ps exposure).
-    let cookie_file: Option<std::path::PathBuf> = if !cookies.is_empty() {
+    let cookie_file: Option<PathBuf> = if !cookies.is_empty() {
         let cf = temp_dir.join("cookies.txt");
         let domain = domain_from_url(url);
         write_netscape_cookie_file(cookies, &domain, &cf)
@@ -106,52 +160,90 @@ fn save_with(
         None
     };
 
-    let mut cmd = Command::new(single_file);
-    cmd.arg(url)
-        .arg(&out_file)
-        .arg(format!("--browser-executable-path={chrome}"))
-        .arg("--browser-headless")
-        .arg("--browser-wait-until=networkidle2")
-        // Extra delay after networkidle2: Cloudflare Fonts injects @font-face
-        // CSS after HTML parse, so the font hook needs more time to see it.
-        .arg("--browser-wait-delay=2000")
-        // Realistic UA: some origins block headless Chrome's default UA string.
-        .arg("--user-agent=Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/137.0.0.0 Safari/537.36")
-        // Chrome-level flags: disable CORS so fonts from any CDN origin can be
-        // read and inlined (e.g. fonts.gstatic.com without ACAO:*).
-        .arg(format!("--browser-args={browser_args}"))
-        // Preserve all CSS: single-file's defaults strip rules it considers
-        // "unused" (breaks CSS nesting) and remove @media blocks that don't
-        // match the capture viewport (breaks responsive layout).
-        .arg("--remove-unused-styles=false")
-        .arg("--remove-alternative-medias=false")
-        // Allow scripts to run during capture so JS-applied classes exist in
-        // the DOM when CSS is evaluated. The user script above strips <script>
-        // elements before serialization so no broken module imports end up in
-        // the saved file.
-        .arg("--block-scripts=false")
-        .arg(format!("--browser-script={}", user_script_path.display()))
-        // Preserve fonts: defaults strip @font-face rules deemed "unused" or
-        // "alternative" (unicode-range subsets), losing CDN-served fonts.
-        .arg("--remove-unused-fonts=false")
-        .arg("--remove-alternative-fonts=false");
-    if let Some(cf) = &cookie_file {
-        cmd.arg(format!("--browser-cookies-file={}", cf.display()));
-    }
-    let spawn_result = cmd
-        .output()
-        .with_context(|| format!("failed to spawn {single_file} process"));
+    let sf_output = match ublock_ext {
+        // ── We own Chrome's lifecycle (uBlock extension mode) ─────────────
+        Some(ext_path) => {
+            let port = allocate_free_port().context("failed to allocate a free TCP port")?;
 
-    // Delete cookie file unconditionally — including on spawn failure —
-    // so secrets are never left in store/temp when the capture fails.
+            let mut chrome_flags = vec![
+                "--headless=new".to_string(),
+                format!("--remote-debugging-port={port}"),
+                format!("--user-data-dir={}", chrome_data_dir.display()),
+                // Load the extension; disable all others so no unexpected ext loads.
+                format!("--load-extension={}", ext_path.display()),
+                format!("--disable-extensions-except={}", ext_path.display()),
+                // Allow cross-origin font inlining (same reason as standalone mode).
+                "--disable-web-security".to_string(),
+                // Realistic viewport so responsive @media rules are preserved.
+                "--window-size=1920,1080".to_string(),
+            ];
+            chrome_flags.extend(extra_chrome_args);
+
+            let mut chrome_child = Command::new(chrome)
+                .args(&chrome_flags)
+                .stdout(std::process::Stdio::null())
+                .stderr(std::process::Stdio::null())
+                .spawn()
+                .with_context(|| format!("failed to spawn Chrome ({chrome})"))?;
+
+            if !wait_for_chrome_ready(port, 10) {
+                let _ = chrome_child.kill();
+                let _ = chrome_child.wait();
+                bail!("Chrome did not become ready on port {port} within 10 s");
+            }
+
+            let out = run_single_file_with_server(
+                url,
+                &out_file,
+                single_file,
+                port,
+                &user_script_path,
+                cookie_file.as_deref(),
+            );
+
+            // Always kill Chrome and reap its exit status, even on single-file failure.
+            let _ = chrome_child.kill();
+            let _ = chrome_child.wait();
+
+            out.with_context(|| format!("failed to spawn single-file ({single_file})"))?
+        }
+
+        // ── single-file manages Chrome (original behaviour) ───────────────
+        None => {
+            let mut chrome_flags = vec![
+                "--disable-web-security".to_string(),
+                format!("--user-data-dir={}", chrome_data_dir.display()),
+                "--window-size=1920,1080".to_string(),
+            ];
+            chrome_flags.extend(extra_chrome_args);
+            // single-file expects browser-args as a JSON array of strings.
+            let quoted: Vec<String> = chrome_flags
+                .iter()
+                .map(|f| format!("\"{}\"", f.replace('\\', "\\\\").replace('"', "\\\"")))
+                .collect();
+            let browser_args = format!("[{}]", quoted.join(","));
+
+            run_single_file_standalone(
+                url,
+                &out_file,
+                single_file,
+                chrome,
+                &browser_args,
+                &user_script_path,
+                cookie_file.as_deref(),
+            )
+            .with_context(|| format!("failed to spawn single-file ({single_file})"))?
+        }
+    };
+
+    // Delete cookie file unconditionally — including on failure — so secrets
+    // are never left in store/temp when the capture fails.
     if let Some(cf) = &cookie_file {
         let _ = std::fs::remove_file(cf);
     }
 
-    let out = spawn_result?;
-
-    if !out.status.success() {
-        let stderr = String::from_utf8_lossy(&out.stderr);
+    if !sf_output.status.success() {
+        let stderr = String::from_utf8_lossy(&sf_output.stderr);
         bail!("single-file failed: {stderr}");
     }
 
@@ -164,12 +256,127 @@ fn save_with(
 
     let title = extract_html_title(&out_file);
     let html_hash = hash_file(&out_file)?;
-    let (favicon_hash, favicon_ext) = extract_and_save_favicon(&out_file, &temp_dir, timestamp)
+    let (favicon_hash, favicon_ext) =
+        extract_and_save_favicon(&out_file, &temp_dir, timestamp)
             .map(|(h, e)| (Some(h), Some(e)))
             .unwrap_or((None, None));
-    Ok(SaveResult { html_hash, title, favicon_hash, favicon_ext })
+
+    Ok(SaveResult {
+        html_hash,
+        title,
+        favicon_hash,
+        favicon_ext,
+        ublock_skipped: false, // overwritten by save() from resolve_ublock_config()
+    })
 }
 
+// ── Chrome helpers ────────────────────────────────────────────────────────────
+
+/// Binds a `TcpListener` to a random OS-assigned port, reads the port number,
+/// then drops the listener.  The tiny TOCTOU window between drop and Chrome
+/// binding is acceptable in practice.
+fn allocate_free_port() -> Result<u16> {
+    let listener =
+        TcpListener::bind("127.0.0.1:0").context("could not bind to a free TCP port")?;
+    Ok(listener.local_addr()?.port())
+}
+
+/// Polls `http://127.0.0.1:{port}/json/version` every 150 ms until Chrome
+/// responds with HTTP 200 or the deadline (timeout_secs) elapses.
+fn wait_for_chrome_ready(port: u16, timeout_secs: u64) -> bool {
+    let url = format!("http://127.0.0.1:{port}/json/version");
+    let client = match reqwest::blocking::Client::builder()
+        .timeout(Duration::from_millis(500))
+        .build()
+    {
+        Ok(c) => c,
+        Err(_) => return false,
+    };
+    let deadline = Instant::now() + Duration::from_secs(timeout_secs);
+    while Instant::now() < deadline {
+        if client
+            .get(&url)
+            .send()
+            .map(|r| r.status().is_success())
+            .unwrap_or(false)
+        {
+            return true;
+        }
+        std::thread::sleep(Duration::from_millis(150));
+    }
+    false
+}
+
+// ── single-file invocation helpers ───────────────────────────────────────────
+
+/// Runs single-file pointing at an already-running Chrome via the DevTools HTTP
+/// API (`--browser-server`).  Chrome was started by the caller, which retains
+/// ownership of the process handle and kills it after this call returns.
+fn run_single_file_with_server(
+    url: &str,
+    out_file: &Path,
+    single_file: &str,
+    port: u16,
+    user_script_path: &Path,
+    cookie_file: Option<&Path>,
+) -> std::io::Result<std::process::Output> {
+    let mut cmd = base_single_file_cmd(url, out_file, single_file, user_script_path, cookie_file);
+    cmd.arg(format!("--browser-server=http://127.0.0.1:{port}"));
+    cmd.output()
+}
+
+/// Runs single-file, letting it launch and manage Chrome itself.
+fn run_single_file_standalone(
+    url: &str,
+    out_file: &Path,
+    single_file: &str,
+    chrome: &str,
+    browser_args: &str,  // JSON array string, e.g. `["--disable-web-security",...]`
+    user_script_path: &Path,
+    cookie_file: Option<&Path>,
+) -> std::io::Result<std::process::Output> {
+    let mut cmd = base_single_file_cmd(url, out_file, single_file, user_script_path, cookie_file);
+    cmd.arg(format!("--browser-executable-path={chrome}"))
+        .arg("--browser-headless")
+        .arg(format!("--browser-args={browser_args}"));
+    cmd.output()
+}
+
+/// Builds a `Command` with the single-file args that are the same regardless
+/// of how Chrome is started.
+fn base_single_file_cmd(
+    url: &str,
+    out_file: &Path,
+    single_file: &str,
+    user_script_path: &Path,
+    cookie_file: Option<&Path>,
+) -> Command {
+    let mut cmd = Command::new(single_file);
+    cmd.arg(url)
+        .arg(out_file)
+        .arg("--browser-wait-until=networkidle2")
+        // Extra delay after networkidle2: Cloudflare Fonts injects @font-face
+        // CSS after HTML parse, so the font hook needs more time to see it.
+        .arg("--browser-wait-delay=2000")
+        // Realistic UA: some origins block headless Chrome's default UA string.
+        .arg("--user-agent=Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/137.0.0.0 Safari/537.36")
+        // Preserve all CSS.
+        .arg("--remove-unused-styles=false")
+        .arg("--remove-alternative-medias=false")
+        // Allow scripts during capture so JS-applied classes exist in the DOM.
+        .arg("--block-scripts=false")
+        .arg(format!("--browser-script={}", user_script_path.display()))
+        // Preserve fonts.
+        .arg("--remove-unused-fonts=false")
+        .arg("--remove-alternative-fonts=false");
+    if let Some(cf) = cookie_file {
+        cmd.arg(format!("--browser-cookies-file={}", cf.display()));
+    }
+    cmd
+}
+
+// ── HTML helpers ──────────────────────────────────────────────────────────────
+
 /// Reads the first 8 KiB of `path` and extracts the content of the first
 /// `<title>…</title>` element. Returns `None` if absent or empty.
 ///
@@ -177,21 +384,17 @@ fn save_with(
 /// lowercasing is byte-length-preserving, so byte offsets derived from the
 /// lowercased buffer are valid indices into the original buffer.
 fn extract_html_title(path: &Path) -> Option<String> {
-    let mut buf = [0u8; 8192];
-    let n = std::fs::File::open(path).ok()?.read(&mut buf).ok()?;
-    // Recover a valid UTF-8 prefix if the 8 KiB boundary falls mid-character.
-    let snippet = match std::str::from_utf8(&buf[..n]) {
-        Ok(s) => s,
-        Err(e) => std::str::from_utf8(&buf[..e.valid_up_to()]).ok()?,
-    };
-    // ASCII-only lowercase: A-Z -> a-z, all other bytes unchanged.
-    // Byte lengths are identical to the original, so offsets are safe to reuse.
-    let lower = snippet.to_ascii_lowercase();
-    let tag_start = lower.find("<title>")?;
-    let content_start = tag_start + 7; // len("<title>") == 7
-    let content_end = content_start + lower[content_start..].find("</title>")?;
-    let title = snippet[content_start..content_end].trim();
-    if title.is_empty() { None } else { Some(title.to_string()) }
+    let mut f = std::fs::File::open(path).ok()?;
+    let mut buf = vec![0u8; 8192];
+    let n = f.read(&mut buf).ok()?;
+    let buf = &buf[..n];
+    let lower = String::from_utf8_lossy(buf).to_ascii_lowercase();
+    let start = lower.find("<title>")? + "<title>".len();
+    let end = lower[start..].find("</title>")? + start;
+    let title = String::from_utf8_lossy(&buf[start..end])
+        .trim()
+        .to_string();
+    if title.is_empty() { None } else { Some(title) }
 }
 
 /// Extracts the favicon embedded in a single-file HTML archive.
@@ -205,71 +408,54 @@ fn extract_and_save_favicon(
     temp_dir: &Path,
     timestamp: &str,
 ) -> Option<(String, String)> {
-    let html = std::fs::read_to_string(html_path).ok()?;
-    let lower = html.to_ascii_lowercase();
+    let content = std::fs::read_to_string(html_path).ok()?;
+    let lower = content.to_ascii_lowercase();
 
-    // Find a <link> tag that has rel="...icon..." AND href="data:image/..."
-    let link_start = {
-        let mut found = None;
-        let mut search = 0;
-        while search < lower.len() {
-            let off = lower[search..].find("<link")?;
-            let abs = search + off;
-            // Find end of this tag, respecting quoted attribute values so that
-            // a '>' inside a data URL does not terminate the tag prematurely.
-            let tag_slice = &lower[abs..];
-            let mut in_q = false;
-            let mut tag_end = None;
-            for (i, c) in tag_slice.char_indices() {
-                match c {
-                    '"' => in_q = !in_q,
-                    '>' if !in_q => { tag_end = Some(i); break; }
-                    _ => {}
-                }
-            }
-            let tag_end = match tag_end { Some(e) => e, None => break };
-            let tag_s = &lower[abs..abs + tag_end];
-            if tag_s.contains("rel=") && tag_s.contains("icon") && tag_s.contains("href=\"data:image") {
-                found = Some(abs);
-                break;
-            }
-            search = abs + tag_end + 1;
-        }
-        found?
+    // Find the first <link …> tag that looks like a favicon with a data: href.
+    let mut search_pos = 0;
+    loop {
+        let tag_start = lower[search_pos..].find("<link")? + search_pos;
+        let tag_end = lower[tag_start..].find('>')? + tag_start;
+        let tag = &lower[tag_start..=tag_end];
+
+        if tag.contains("icon") {
+            // Look for href="data:image/...;base64,..."
+            if let Some(href_pos) = tag.find("href=") {
+                let after_href = &content[tag_start + href_pos + 5..];
+                let (quote, after_quote) = if after_href.starts_with('"') {
+                    ('"', &after_href[1..])
+                } else if after_href.starts_with('\'') {
+                    ('\'', &after_href[1..])
+                } else {
+                    search_pos = tag_end + 1;
+                    continue;
                 };
-
-    // Extract href value from the original HTML (byte positions match because
-    // to_ascii_lowercase is byte-length-preserving).
-    let tag_lower = &lower[link_start..];
-    let href_off = tag_lower.find("href=\"")?;
-    let value_start = link_start + href_off + 6; // past href="
-    let value_end = html[value_start..].find('"')?;
-    let data_url = &html[value_start..value_start + value_end];
-
-    // Parse data:<mime>;base64,<payload>
-    let rest = data_url.strip_prefix("data:")?;
-    let comma = rest.find(',')?;
-    let meta = &rest[..comma];
-    let b64 = &rest[comma + 1..];
-    if !meta.to_ascii_lowercase().contains("base64") {
-        return None;
+                let value_end = after_quote.find(quote)?;
+                let href_value = &after_quote[..value_end];
+                if let Some(b64_start) = href_value.to_ascii_lowercase().find(";base64,") {
+                    let mime_part = &href_value[5..b64_start]; // skip "data:"
+                    let ext = mime_to_favicon_ext(mime_part)?;
+                    let b64_data = &href_value[b64_start + 8..];
+                    let bytes = B64.decode(b64_data).ok()?;
+                    let out_path = temp_dir.join(format!("{timestamp}.favicon{ext}"));
+                    std::fs::write(&out_path, &bytes).ok()?;
+                    let hash = hash_file(&out_path).ok()?;
+                    return Some((hash, ext.to_string()));
+                }
+            }
         }
-    let mime = meta.split(';').next()?.trim().to_ascii_lowercase();
-    let ext = mime_to_favicon_ext(&mime)?;
 
-    let bytes = B64.decode(b64.trim()).ok()?;
-    let out = temp_dir.join(format!("{timestamp}.favicon{ext}"));
-    std::fs::write(&out, &bytes).ok()?;
-    hash_file(&out).ok().map(|h| (h, ext.to_string()))
+        search_pos = tag_end + 1;
+    }
 }
 
 fn mime_to_favicon_ext(mime: &str) -> Option<&'static str> {
-    match mime {
+    match mime.to_ascii_lowercase().trim() {
         "image/x-icon" | "image/vnd.microsoft.icon" => Some(".ico"),
         "image/png"  => Some(".png"),
-        "image/svg+xml" => Some(".svg"),
         "image/jpeg" => Some(".jpg"),
         "image/gif"  => Some(".gif"),
+        "image/svg+xml" => Some(".svg"),
         "image/webp" => Some(".webp"),
         _ => None,
     }
@@ -323,6 +509,7 @@ mod tests {
             "/nonexistent/single-file",
             "chromium",
             &HashMap::new(),
+            None, // no ublock ext
         );
         let err = result.unwrap_err();
         let msg = format!("{err:#}");
@@ -331,4 +518,19 @@ mod tests {
             "unexpected error: {msg}"
         );
     }
+
+    #[test]
+    fn resolve_ublock_config_disabled_when_false() {
+        // Can't mutate env vars safely in parallel tests; test the logic directly
+        // by verifying the env-var parsing branch we care about.
+        let enabled = "false";
+        let is_disabled =
+            enabled.eq_ignore_ascii_case("false") || enabled == "0";
+        assert!(is_disabled);
+
+        let enabled = "0";
+        let is_disabled =
+            enabled.eq_ignore_ascii_case("false") || enabled == "0";
+        assert!(is_disabled);
+    }
 }

crates/archivr-server/src/routes.rs

@@ -756,15 +756,21 @@ async fn capture_handler(
                 return;
             }
         };
-        database::update_capture_job_status(&conn, &job_uid_bg, "running", None, None).ok();
+        database::update_capture_job_status(&conn, &job_uid_bg, "running", None, None, None).ok();
         match capture::perform_capture(&archive_paths, &locator, Some(&archive_id_bg), quality.as_deref(), &capture_config) {
             Ok(result) => {
+                let notes = if result.ublock_skipped {
+                    Some(r#"{"ublock_skipped":true}"#)
+                } else {
+                    None
+                };
                 database::update_capture_job_status(
                     &conn,
                     &job_uid_bg,
                     "completed",
                     Some(&result.run_uid),
                     None,
+                    notes,
                 )
                 .ok();
             }
@@ -775,6 +781,7 @@ async fn capture_handler(
                     "failed",
                     None,
                     Some(&format!("{e:#}")),
+                    None,
                 )
                 .ok();
             }

crates/archivr-server/static/index.html

@@ -4,8 +4,8 @@
     <meta charset="utf-8" />
     <meta name="viewport" content="width=device-width, initial-scale=1" />
     <title>Archivr</title>
-    <script type="module" crossorigin src="/assets/index-CHDuICqH.js"></script>
-    <link rel="stylesheet" crossorigin href="/assets/index-BW0QKHXE.css">
+    <script type="module" crossorigin src="/assets/index-BpXoSc97.js"></script>
+    <link rel="stylesheet" crossorigin href="/assets/index-BGSI2WEx.css">
   </head>
   <body>
     <div id="root"></div>

flake.nix

@@ -60,6 +60,22 @@
           tweetPython = pkgs.python312.withPackages (ps: [
             twitterApiClient
           ]);
+          # uBlock Origin Lite (MV3) — unpacked Chromium extension for headless ad-blocking.
+          # Fetched from the uBOL-home GitHub releases; update version + hash together.
+          ublockLite = pkgs.stdenv.mkDerivation {
+            pname = "ublock-origin-lite";
+            version = "2026.705.2152";
+            src = pkgs.fetchurl {
+              url = "https://github.com/uBlockOrigin/uBOL-home/releases/download/2026.705.2152/uBOLite_2026.705.2152.chromium.zip";
+              hash = "sha256-4TbvDYbkOkDuVK17TeAbLDBcgf9O6f/vh2buGbLu4XQ=";
+            };
+            nativeBuildInputs = [ pkgs.unzip ];
+            sourceRoot = ".";
+            installPhase = ''
+              mkdir -p $out
+              cp -r . $out/
+            '';
+          };
           version = "0.1.0";
           src = pkgs.lib.cleanSource ./.;
           cargoLock = {
@@ -123,6 +139,7 @@
                 ${lib.optionalString pkgs.stdenv.isLinux "--set ARCHIVR_CHROME ${pkgs.chromium}/bin/chromium"} \
                 --set ARCHIVR_TWEET_PYTHON ${tweetPython}/bin/python3 \
                 --set ARCHIVR_TWEET_SCRAPER $out/libexec/archivr/scrape_user_tweet_contents.py \
+                --set ARCHIVR_UBLOCK_EXT ${ublockLite} \
                 --prefix PATH : ${
                   lib.makeBinPath ([
                     pkgs.yt-dlp
@@ -149,7 +166,8 @@
                 --set ARCHIVR_SINGLE_FILE ${pkgs.single-file-cli}/bin/single-file \
                 ${lib.optionalString pkgs.stdenv.isLinux "--set ARCHIVR_CHROME ${pkgs.chromium}/bin/chromium"} \
                 --set ARCHIVR_TWEET_PYTHON ${tweetPython}/bin/python3 \
-                --set ARCHIVR_TWEET_SCRAPER $out/libexec/archivr-server/scrape_user_tweet_contents.py
+                --set ARCHIVR_TWEET_SCRAPER $out/libexec/archivr-server/scrape_user_tweet_contents.py \
+                --set ARCHIVR_UBLOCK_EXT ${ublockLite}
             '';
           };
           archivr-all = pkgs.symlinkJoin {

frontend/src/App.jsx

@@ -85,6 +85,9 @@ export default function App() {
 
   const [toasts, setToasts] = useState([])
   const toastIdRef = useRef(0)
+  const [ublockWarningIgnored, setUblockWarningIgnored] = useState(
+    () => sessionStorage.getItem('ublockWarningIgnored') === 'true'
+  )
 
   const humanizeTags = currentUser?.humanize_slugs ?? false;
 
@@ -238,15 +241,22 @@ export default function App() {
     ])
   }, [archiveId, searchQuery, tagFilter, loadEntries])
 
-  const handleToast = useCallback((errorText, locator) => {
+  const handleToast = useCallback((text, locator, type = 'error') => {
+    if (type === 'warning' && ublockWarningIgnored) return
     const id = ++toastIdRef.current
-    setToasts(prev => [...prev, { id, errorText, locator }])
-  }, [])
+    setToasts(prev => [...prev, { id, text, locator, type }])
+  }, [ublockWarningIgnored])
 
   const handleDismissToast = useCallback((id) => {
     setToasts(prev => prev.filter(t => t.id !== id))
   }, [])
 
+  const handleIgnoreUblock = useCallback(() => {
+    sessionStorage.setItem('ublockWarningIgnored', 'true')
+    setUblockWarningIgnored(true)
+    setToasts(prev => prev.filter(t => t.type !== 'warning'))
+  }, [])
+
   if (authState === 'loading') return <div className="auth-loading">Loading\u2026</div>;
   if (authState === 'setup')   return <SetupPage onComplete={() => setAuthState('login')} />;
   if (authState === 'login')   return <LoginPage onLogin={user => { setCurrentUser(user); setAuthState('authenticated'); }} />;
@@ -347,7 +357,7 @@ export default function App() {
           onCaptured={handleCaptured}
           onToast={handleToast}
         />
-        <ToastStack toasts={toasts} onDismiss={handleDismissToast} />
+        <ToastStack toasts={toasts} onDismiss={handleDismissToast} onIgnoreUblock={handleIgnoreUblock} />
       </>
     </AuthContext.Provider>
   )

frontend/src/components/CaptureDialog.jsx

@@ -186,6 +186,13 @@ export default function CaptureDialog({ open, archiveId, onClose, onCaptured, on
             })
           }, 1400)
           onCapturedRef.current()
+          // Warn if uBlock was requested but the extension wasn't available
+          try {
+            const notes = updated.notes_json ? JSON.parse(updated.notes_json) : null
+            if (notes?.ublock_skipped) {
+              onToastRef.current(null, locator, 'warning')
+            }
+          } catch {}
         } else if (updated.status === 'failed') {
           clearInterval(pollIntervals.current.get(jobUid))
           pollIntervals.current.delete(jobUid)

frontend/src/components/ToastStack.jsx

@@ -1,22 +1,23 @@
 import { useState, useEffect } from 'react'
 
-const TOAST_TTL = 7000 // ms before auto-dismiss; paused when error is expanded
+const TOAST_TTL = 7000 // ms before auto-dismiss; paused when detail is expanded
 
-export default function ToastStack({ toasts, onDismiss }) {
+export default function ToastStack({ toasts, onDismiss, onIgnoreUblock }) {
   if (!toasts.length) return null
   return (
     <div className="toast-stack" role="log" aria-live="polite" aria-label="Notifications">
       {toasts.map(t => (
-        <Toast key={t.id} toast={t} onDismiss={onDismiss} />
+        <Toast key={t.id} toast={t} onDismiss={onDismiss} onIgnoreUblock={onIgnoreUblock} />
       ))}
     </div>
   )
 }
 
-function Toast({ toast, onDismiss }) {
+function Toast({ toast, onDismiss, onIgnoreUblock }) {
   const [expanded, setExpanded] = useState(false)
+  const isWarning = toast.type === 'warning'
 
-  // Auto-dismiss after TTL; paused while error detail is expanded
+  // Auto-dismiss after TTL; paused while detail is expanded
   useEffect(() => {
     if (expanded) return
     const timer = setTimeout(() => onDismiss(toast.id), TOAST_TTL)
@@ -27,6 +28,50 @@ function Toast({ toast, onDismiss }) {
     ? (toast.locator.length > 48 ? toast.locator.slice(0, 45) + '\u2026' : toast.locator)
     : null
 
+  if (isWarning) {
+    return (
+      <div className="toast toast--warning" role="alert" aria-atomic="true">
+        <div className="toast-top">
+          <span className="toast-icon" aria-hidden="true">⚠</span>
+          <div className="toast-body">
+            <span className="toast-headline">Ad-blocking unavailable</span>
+            {short && <span className="toast-locator">{short}</span>}
+          </div>
+          <div className="toast-btns">
+            <button
+              type="button"
+              className="toast-view-btn"
+              onClick={() => setExpanded(v => !v)}
+              aria-expanded={expanded}
+            >
+              {expanded ? 'Hide' : 'Details'}
+            </button>
+            <button
+              type="button"
+              className="toast-view-btn toast-ignore-btn"
+              onClick={() => { onIgnoreUblock?.(); onDismiss(toast.id) }}
+            >
+              Ignore
+            </button>
+            <button
+              type="button"
+              className="toast-dismiss"
+              onClick={() => onDismiss(toast.id)}
+              aria-label="Dismiss"
+            >
+              ×
+            </button>
+          </div>
+        </div>
+        {expanded && (
+          <p className="toast-warning-detail">
+            {toast.text || 'ARCHIVR_UBLOCK=true but ARCHIVR_UBLOCK_EXT is not set or the path is invalid. The page was captured without ad-blocking. Set ARCHIVR_UBLOCK_EXT to the unpacked uBlock Origin Lite extension directory to enable ad-blocking, or set ARCHIVR_UBLOCK=false to silence this warning.'}
+          </p>
+        )}
+      </div>
+    )
+  }
+
   return (
     <div className="toast toast--error" role="alert" aria-atomic="true">
       <div className="toast-top">
@@ -36,7 +81,7 @@ function Toast({ toast, onDismiss }) {
           {short && <span className="toast-locator">{short}</span>}
         </div>
         <div className="toast-btns">
-          {toast.errorText && (
+          {toast.text && (
             <button
               type="button"
               className="toast-view-btn"
@@ -55,8 +100,8 @@ function Toast({ toast, onDismiss }) {
           </button>
         </div>
       </div>
-      {expanded && toast.errorText && (
-        <pre className="toast-error-detail">{toast.errorText}</pre>
+      {expanded && toast.text && (
+        <pre className="toast-error-detail">{toast.text}</pre>
       )}
     </div>
   )

frontend/src/styles.css

@@ -812,6 +812,7 @@ select {
   to   { opacity: 1; transform: translateY(0) scale(1); }
 }
 .toast--error { border-left: 3px solid var(--accent); }
+.toast--warning { border-left: 3px solid #e8a000; }
 
 .toast-top {
   display: flex;
@@ -892,6 +893,18 @@ select {
   max-height: 200px;
   overflow-y: auto;
 }
+.toast-warning-detail {
+  margin: 0;
+  padding: 10px 14px 12px;
+  font-size: 12px;
+  line-height: 1.55;
+  color: var(--muted);
+  background: var(--paper-2);
+  border-top: 1px solid var(--line);
+  white-space: pre-wrap;
+  word-break: break-word;
+}
+.toast-ignore-btn { color: var(--muted); }
 
 /* ── Utility ─────────────────────────────────────────────────────────────── */
 .muted { color: var(--muted); }