diff --git a/docs/web-ui/dashboard.md b/docs/web-ui/dashboard.md
index 3baa5a59..fef65252 100644
--- a/docs/web-ui/dashboard.md
+++ b/docs/web-ui/dashboard.md
@@ -153,22 +153,6 @@ age + claude-creds badge). Two actions: `⊕ R3V1V3` (queues a
 Spawn approval; existing state is reused), `PURG3` (wipes
 state + applied dirs; `POST /purge-tombstone/{name}`).
 
-**C0NT41N3R L04D** — live CPU + memory per agent container, read
-straight from cgroup v2 on the host (`cpu.stat`, `memory.current`,
-`memory.peak`, `memory.max` under
-`/sys/fs/cgroup/machine.slice/machine-h\x2d<name>.scope/`). CPU is a
-host-normalised percentage (0..100 across all cores) sampled over a
-short (~200 ms) two-read interval; memory shows current + peak with a
-bar against the `memory.max` quota. Backed by
-`GET /api/container-resources` (`container_stats.rs`), which reads the
-files read-only (world-readable; no `hive-priv`) and skips agents whose
-scope dir is absent (= not running). Pull-only: `tabs.js` polls every
-5 s **only while the SYST3M tab is active** (CPU needs a fresh sample
-each refresh), and stops on tab change. Network is intentionally
-omitted — agents share the host netns, so there is no per-container net
-counter (per-agent network needs the netns-isolation roadmap in
-`docs/network.md`).
-
 ## P3RM1SS10NS tab
 
 Per-agent permission configuration. Two sections, each rendered as a
@@ -854,12 +838,6 @@ that's a browser-level decision, not ours.
   zero-turn dbs); returns swarm totals, a busiest-first per-agent
   rollup, swarm model mix, and a labelled `est_cost_usd`. Window
   defaults to `24h`.
-- `GET /api/container-resources` — live per-agent-container CPU +
-  memory from cgroup v2 (SYST3M › C0NT41N3R L04D panel). Returns one
-  row per running agent (`name`, `cpu_pct`, `mem_current_bytes`,
-  `mem_peak_bytes`, `mem_max_bytes`); samples CPU over ~200 ms so the
-  call briefly awaits. Skips non-running agents (no scope dir). No
-  network field — agents share the host netns.
 - `POST /cancel-reminder/{id}` — hard-delete a pending reminder.
 - `POST /retry-reminder/{id}` — re-arm a reminder whose delivery
   failed (clears the failure state so the scheduler retries).
diff --git a/frontend/packages/dashboard/src/dashboard.css b/frontend/packages/dashboard/src/dashboard.css
index 5d1f3294..4c707dbd 100644
--- a/frontend/packages/dashboard/src/dashboard.css
+++ b/frontend/packages/dashboard/src/dashboard.css
@@ -1609,29 +1609,3 @@ body.dashboard-shell.has-selection { padding-bottom: 4.5em; }
   color: var(--muted);
   font-variant-numeric: tabular-nums;
 }
-
-/* ─── SYST3M › C0NT41N3R L04D: live cgroup cpu/mem ─────────────────────────────
-   Reuses the `.hive-stats-table` styling from the ST4TS tab; only the
-   inline meter bar is new. */
-.cload-meter {
-  display: inline-block;
-  width: 6em;
-  height: 10px;
-  vertical-align: middle;
-  margin-left: 6px;
-  background: var(--bg-elev);
-  border: 1px solid var(--border);
-  border-radius: 3px;
-  overflow: hidden;
-}
-.cload-meter .fill {
-  display: block;
-  height: 100%;
-  background: var(--green);
-}
-.cload-meter .fill.warn {
-  background: var(--amber);
-}
-.cload-meter .fill.hot {
-  background: var(--red);
-}
diff --git a/frontend/packages/dashboard/src/index.html b/frontend/packages/dashboard/src/index.html
index 2b3f8bc1..e81b52c0 100644
--- a/frontend/packages/dashboard/src/index.html
+++ b/frontend/packages/dashboard/src/index.html
@@ -191,16 +191,6 @@
         <p class="meta">loading…</p>
       </div>
 
-      <!-- C0NT41N3R L04D: live cpu + memory per agent container, read
-           from cgroup v2 on the host. Polled only while this tab is
-           active (cpu needs a short two-sample read each refresh). -->
-      <h2>◆ C0NT41N3R L04D ◆</h2>
-      <div class="divider">══════════════════════════════════════════════════════════════</div>
-      <p class="meta">live cpu + memory per agent container, from cgroup v2 on the host. cpu is % of total host capacity (all cores), sampled over ~200ms each refresh; polled every 5s while this tab is open. network is omitted on purpose — agents share the host netns, so there is no per-container counter (see <code>docs/network.md</code>).</p>
-      <div id="container-load-section">
-        <p class="meta">loading…</p>
-      </div>
-
     </section>
 
     <!-- P3RM1SS10NS: per-agent capability grants + tool-group
diff --git a/frontend/packages/dashboard/src/tabs.js b/frontend/packages/dashboard/src/tabs.js
index b37bc7cb..3e22ee8e 100644
--- a/frontend/packages/dashboard/src/tabs.js
+++ b/frontend/packages/dashboard/src/tabs.js
@@ -4076,9 +4076,6 @@ window.marked = marked;
     }
     // ST4TS: hive-wide rollup is a pull (no SSE) — fetch on activation.
     if (target === 'stats') { refreshHiveStats(); }
-    // SYST3M › C0NT41N3R L04D: live cgroup poll only while the tab is
-    // open (cpu needs a short two-sample read each refresh).
-    if (target === 'system') { startContainerLoadPolling(); } else { stopContainerLoadPolling(); }
   }
   // ─── tabbar overflow menu ────────────────────────────────────────────────
   // Tabs with `data-overflow="default"` (LOGS, SETTINGS) always live in
@@ -4233,100 +4230,6 @@ window.marked = marked;
   }
   bindHiveStatsWindows();
 
-  // ─── SYST3M › C0NT41N3R L04D: live per-container cgroup cpu/mem ────────────
-  // Pull-only, polled at 5s ONLY while the SYST3M tab is active (cpu is a
-  // short two-sample read on the server). Data from /api/container-resources.
-  let containerLoadTimer = null;
-
-  function cloadFmtBytes(n) {
-    if (!Number.isFinite(n) || n <= 0) return '0';
-    const u = ['B', 'KiB', 'MiB', 'GiB', 'TiB'];
-    let i = 0; let v = n;
-    while (v >= 1024 && i < u.length - 1) { v /= 1024; i += 1; }
-    return v.toFixed(v < 10 && i > 0 ? 1 : 0) + ' ' + u[i];
-  }
-  function cloadMeter(pct) {
-    const p = Math.max(0, Math.min(100, pct));
-    const cls = p >= 90 ? 'hot' : (p >= 70 ? 'warn' : '');
-    const m = document.createElement('span');
-    m.className = 'cload-meter';
-    m.title = p.toFixed(0) + '%';
-    const f = document.createElement('span');
-    f.className = 'fill' + (cls ? ' ' + cls : '');
-    f.style.width = p + '%';
-    m.append(f);
-    return m;
-  }
-
-  function renderContainerLoad(rows) {
-    const root = $('container-load-section');
-    if (!root) return;
-    if (!Array.isArray(rows) || rows.length === 0) {
-      root.replaceChildren();
-      const p = document.createElement('p'); p.className = 'meta';
-      p.textContent = 'no running agent containers'; root.append(p);
-      return;
-    }
-    const table = document.createElement('table');
-    table.className = 'hive-stats-table';
-    table.innerHTML = '<thead><tr><th>agent</th><th>cpu</th><th>memory</th>'
-      + '<th>peak</th><th>limit</th></tr></thead>';
-    const tb = document.createElement('tbody');
-    for (const r of rows) {
-      const tr = document.createElement('tr');
-
-      const name = document.createElement('td'); name.textContent = r.name; tr.append(name);
-
-      const cpu = document.createElement('td'); cpu.className = 'num';
-      cpu.append((Number(r.cpu_pct) || 0).toFixed(1) + '%', cloadMeter(Number(r.cpu_pct) || 0));
-      tr.append(cpu);
-
-      const mem = document.createElement('td'); mem.className = 'num';
-      const memCur = Number(r.mem_current_bytes) || 0;
-      if (r.mem_max_bytes) {
-        mem.append(cloadFmtBytes(memCur), cloadMeter(100 * memCur / r.mem_max_bytes));
-      } else {
-        mem.textContent = cloadFmtBytes(memCur);
-      }
-      tr.append(mem);
-
-      const peak = document.createElement('td'); peak.className = 'num';
-      peak.textContent = r.mem_peak_bytes ? cloadFmtBytes(Number(r.mem_peak_bytes)) : '—';
-      tr.append(peak);
-
-      const lim = document.createElement('td'); lim.className = 'num';
-      lim.textContent = r.mem_max_bytes ? cloadFmtBytes(Number(r.mem_max_bytes)) : '∞';
-      tr.append(lim);
-
-      tb.append(tr);
-    }
-    table.append(tb);
-    root.replaceChildren(table);
-  }
-
-  async function refreshContainerLoad() {
-    try {
-      const resp = await fetch('/api/container-resources');
-      if (!resp.ok) throw new Error('http ' + resp.status);
-      renderContainerLoad(await resp.json());
-    } catch (e) {
-      const root = $('container-load-section');
-      if (root) {
-        root.replaceChildren();
-        const p = document.createElement('p'); p.className = 'meta';
-        p.textContent = 'container load fetch failed: ' + e; root.append(p);
-      }
-    }
-  }
-  function startContainerLoadPolling() {
-    refreshContainerLoad();
-    if (containerLoadTimer) return;
-    containerLoadTimer = setInterval(refreshContainerLoad, 5000);
-  }
-  function stopContainerLoadPolling() {
-    if (containerLoadTimer) { clearInterval(containerLoadTimer); containerLoadTimer = null; }
-  }
-
   function syncTabFromHash() {
     const h = (window.location.hash || '#swarm').replace(/^#/, '');
     activateTab(h);
diff --git a/hive-c0re/src/container_stats.rs b/hive-c0re/src/container_stats.rs
deleted file mode 100644
index ea488da4..00000000
--- a/hive-c0re/src/container_stats.rs
+++ /dev/null
@@ -1,168 +0,0 @@
-//! Live per-agent-container resource load (CPU + memory) for the
-//! dashboard. Reads cgroup v2 stat files for each running agent
-//! machine directly — same privsep-clean posture as the turn-stats
-//! sqlite reads (`cpu.stat` / `memory.*` are world-readable; no
-//! `hive-priv` needed). Runs host-side in hive-c0re, where
-//! `/sys/fs/cgroup/machine.slice/` holds the nspawn machine scopes.
-//!
-//! No network: agents share the host network namespace
-//! (`privateNetwork = false`), so there is no per-container net
-//! counter to read. Per-agent network only becomes meaningful with the
-//! netns-isolation roadmap (`docs/network.md`).
-//!
-//! CPU is cumulative (`usage_usec` is monotonic), so a single read is
-//! meaningless — we sample every machine's counter, sleep one short
-//! interval, sample again, and divide the delta by `interval × nproc`
-//! to get a host-normalised percentage (0..100 across all cores).
-
-use std::path::PathBuf;
-use std::time::Duration;
-
-use serde::Serialize;
-use tokio::time::sleep;
-
-use crate::coordinator::Coordinator;
-
-/// Sampling interval for the two CPU reads. Short enough to keep the
-/// endpoint snappy, long enough that the delta isn't dominated by read
-/// jitter.
-const CPU_SAMPLE: Duration = Duration::from_millis(200);
-
-const MACHINE_SLICE: &str = "/sys/fs/cgroup/machine.slice";
-
-#[derive(Debug, Serialize)]
-pub struct ContainerResource {
-    /// Agent name (without the `h-` machine prefix).
-    pub name: String,
-    /// Host-normalised CPU usage over the sample interval, as a
-    /// percentage of total host CPU (0..100 across all cores).
-    pub cpu_pct: f64,
-    /// Current memory usage (`memory.current`), bytes.
-    pub mem_current_bytes: u64,
-    /// High-water memory usage since container start (`memory.peak`),
-    /// bytes. `None` if the kernel doesn't expose `memory.peak`.
-    pub mem_peak_bytes: Option<u64>,
-    /// Memory ceiling (`memory.max`), bytes. `None` when unlimited
-    /// (the file reads `max`).
-    pub mem_max_bytes: Option<u64>,
-}
-
-/// systemd escapes the machine name in the cgroup scope dir
-/// (`machine-<escaped>.scope`). For our machine names — `h-<agent>`,
-/// agent ∈ `[a-z0-9_-]` — the only character systemd escapes is `-`,
-/// which becomes `\x2d` (verified against `systemd-escape`).
-fn escape_machine(machine: &str) -> String {
-    machine.replace('-', "\\x2d")
-}
-
-fn scope_dir(machine: &str) -> PathBuf {
-    PathBuf::from(MACHINE_SLICE).join(format!("machine-{}.scope", escape_machine(machine)))
-}
-
-/// Read a single unsigned integer from a one-line cgroup file.
-fn read_u64(path: &std::path::Path) -> Option<u64> {
-    std::fs::read_to_string(path)
-        .ok()?
-        .trim()
-        .parse::<u64>()
-        .ok()
-}
-
-/// `memory.max` reads `max` when there's no limit — map that to `None`.
-fn read_mem_max(path: &std::path::Path) -> Option<u64> {
-    let s = std::fs::read_to_string(path).ok()?;
-    let s = s.trim();
-    if s == "max" {
-        None
-    } else {
-        s.parse::<u64>().ok()
-    }
-}
-
-/// Parse `usage_usec` (cumulative CPU time, microseconds) from a
-/// scope's `cpu.stat`.
-fn read_usage_usec(dir: &std::path::Path) -> Option<u64> {
-    let text = std::fs::read_to_string(dir.join("cpu.stat")).ok()?;
-    for line in text.lines() {
-        if let Some(rest) = line.strip_prefix("usage_usec ") {
-            return rest.trim().parse::<u64>().ok();
-        }
-    }
-    None
-}
-
-/// Total online host CPUs, for normalising CPU% to "% of total host".
-/// Reads `/sys/devices/system/cpu/online` (a comma-separated list of
-/// ranges, e.g. `0-19`) rather than `available_parallelism()`, which
-/// reflects the hive-core process's CPU affinity — if hive-core is ever
-/// affinity-pinned to a subset, that would under-count the denominator
-/// and inflate the percentage. Falls back to the process count, then 1.
-fn host_nproc() -> usize {
-    fn from_sysfs() -> Option<usize> {
-        let s = std::fs::read_to_string("/sys/devices/system/cpu/online").ok()?;
-        let mut count = 0usize;
-        for part in s.trim().split(',') {
-            let part = part.trim();
-            if part.is_empty() {
-                continue;
-            }
-            if let Some((a, b)) = part.split_once('-') {
-                let a: usize = a.trim().parse().ok()?;
-                let b: usize = b.trim().parse().ok()?;
-                count += b.saturating_sub(a) + 1;
-            } else {
-                part.parse::<usize>().ok()?;
-                count += 1;
-            }
-        }
-        (count > 0).then_some(count)
-    }
-    from_sysfs()
-        .or_else(|| std::thread::available_parallelism().ok().map(usize::from))
-        .unwrap_or(1)
-}
-
-/// Sample CPU + memory for every running agent container. Best-effort:
-/// agents whose scope dir is absent (not running) or unreadable are
-/// skipped, never fatal. Sorted by name for a stable display order.
-pub async fn gather() -> Vec<ContainerResource> {
-    // Machines that currently have a cgroup scope (= running). Agent
-    // machine name is `h-<state-dir name>`.
-    let candidates: Vec<(String, PathBuf)> = Coordinator::kept_state_names()
-        .into_iter()
-        .filter_map(|name| {
-            let dir = scope_dir(&format!("h-{name}"));
-            dir.join("cpu.stat").exists().then_some((name, dir))
-        })
-        .collect();
-
-    // First CPU sample for all, then one shared sleep, then the second
-    // — so N agents cost one interval, not N.
-    let t0: Vec<Option<u64>> = candidates.iter().map(|(_, d)| read_usage_usec(d)).collect();
-    sleep(CPU_SAMPLE).await;
-
-    #[allow(clippy::cast_precision_loss)]
-    let nproc = host_nproc() as f64;
-    #[allow(clippy::cast_precision_loss)]
-    let interval_usec = CPU_SAMPLE.as_micros() as f64;
-
-    let mut out: Vec<ContainerResource> = Vec::with_capacity(candidates.len());
-    for (i, (name, dir)) in candidates.iter().enumerate() {
-        let cpu_pct = match (t0[i], read_usage_usec(dir)) {
-            (Some(a), Some(b)) => {
-                #[allow(clippy::cast_precision_loss)]
-                let delta = b.saturating_sub(a) as f64;
-                (delta / (interval_usec * nproc)) * 100.0
-            }
-            _ => 0.0,
-        };
-        out.push(ContainerResource {
-            name: name.clone(),
-            cpu_pct,
-            mem_current_bytes: read_u64(&dir.join("memory.current")).unwrap_or(0),
-            mem_peak_bytes: read_u64(&dir.join("memory.peak")),
-            mem_max_bytes: read_mem_max(&dir.join("memory.max")),
-        });
-    }
-    out
-}
diff --git a/hive-c0re/src/dashboard.rs b/hive-c0re/src/dashboard.rs
index 1ca26b7b..3739027f 100644
--- a/hive-c0re/src/dashboard.rs
+++ b/hive-c0re/src/dashboard.rs
@@ -68,7 +68,6 @@ pub async fn serve(port: u16, coord: Arc<Coordinator>) -> Result<()> {
         .route("/api/state-file", get(get_state_file))
         .route("/api/reminders", get(api_reminders))
         .route("/api/stats-hive", get(api_stats_hive))
-        .route("/api/container-resources", get(api_container_resources))
         .route("/api/build-logs", get(get_build_logs_all))
         .route("/api/build-logs/{agent}", get(get_build_logs_agent))
         .route("/api/build-logs/id/{id}", get(get_build_log_full))
@@ -1749,12 +1748,6 @@ async fn api_stats_hive(axum::extract::Query(q): axum::extract::Query<StatsHiveQ
     axum::Json(crate::hive_stats::hive_snapshot(window)).into_response()
 }
 
-/// Live per-agent-container CPU + memory load from cgroup v2. Samples
-/// CPU over a short interval (~200 ms), so this call briefly awaits.
-async fn api_container_resources() -> Response {
-    axum::Json(crate::container_stats::gather().await).into_response()
-}
-
 #[derive(Deserialize)]
 struct BuildLogsQuery {
     /// Maximum number of rows to return. Capped server-side at 50
diff --git a/hive-c0re/src/lib.rs b/hive-c0re/src/lib.rs
index 1bba3cad..dce299a7 100644
--- a/hive-c0re/src/lib.rs
+++ b/hive-c0re/src/lib.rs
@@ -23,7 +23,6 @@ pub mod broker;
 pub mod build_logs;
 pub mod capabilities;
 pub mod client;
-pub mod container_stats;
 pub mod container_view;
 pub mod coordinator;
 pub mod crash_watch;