fix(#2289): wire sync_agent failures to dashboard warning banner

sync_agent() now returns bool (false if any step fails). ensure_all()
collects the names of agents whose sync failed and raises a single
set_boot_warning with the aggregated list:

  forge: per-agent sync failed for: alice, bob (see journal for per-step
  detail)

The static_kind() leak is already used for per-org boot warnings in the
same file — the leak is bounded (one per hive-c0re boot, not per request)
so reusing it here is appropriate.

The rebuild call site in job_queue/exec.rs discards the bool return and
keeps its existing tracing::warn! lines, which is the right separation:
rebuilds are their own retry loop and don't need to post a persistent boot
warning.

frontend/packages/dashboard/src/logs.js

@@ -45,8 +45,15 @@ import { createTabStrip } from '@hive/shared/tabs.js';
     return 'fetched ' + (ageSecs < 5 ? 'just now' : fmtAgeSecs(ageSecs) + ' ago');
   }
 
-  // Populate agent selector from /api/state, then honour any `?agent=` /
-  // `?unit=` URL params (used by the per-agent ⋮ menu's deep-link).
+  // Names from `state.infra_containers` (hive-ci / hive-forge / hive-gateway
+  // / hive-matrix) — kept in sync with the fetched list so fetchAgent() can
+  // tell an infra container apart from an agent and skip the (inapplicable)
+  // unit filter for it.
+  let infraContainerNames = new Set();
+
+  // Populate agent selector from /api/state (agents, then the four infra
+  // containers in their own optgroup), then honour any `?agent=` / `?unit=`
+  // URL params (used by the per-agent ⋮ menu's deep-link).
   async function loadAgentList() {
     try {
       const resp = await fetch('/api/state');
@@ -58,6 +65,14 @@ import { createTabStrip } from '@hive/shared/tabs.js';
       for (const c of (state.containers || [])) {
         agentSelect.append(el('option', { value: c.name }, c.name));
       }
+      infraContainerNames = new Set((state.infra_containers || []).map((c) => c.name));
+      if (infraContainerNames.size) {
+        const infraGroup = el('optgroup', { label: 'infra' });
+        for (const name of infraContainerNames) {
+          infraGroup.append(el('option', { value: name }, name));
+        }
+        agentSelect.append(infraGroup);
+      }
       // Deep-link: honour ?agent= and ?unit= URL params.
       const urlAgent = new URLSearchParams(location.search).get('agent');
       const urlUnit  = new URLSearchParams(location.search).get('unit');
@@ -70,12 +85,24 @@ import { createTabStrip } from '@hive/shared/tabs.js';
               .some((o) => o.value === urlUnit);
             if (unitFound) agentUnitSelect.value = urlUnit;
           }
+          syncUnitSelectForSelection();
           fetchAgent();
         }
       }
     } catch { /**/ }
   }
 
+  // Infra containers don't run the per-agent hive daemons, so the unit
+  // filter is meaningless for them — disable the selector and always fetch
+  // the full machine journal to avoid a picked unit silently doing nothing.
+  function syncUnitSelectForSelection() {
+    if (!agentUnitSelect) return;
+    const isInfra = infraContainerNames.has(agentSelect ? agentSelect.value : '');
+    agentUnitSelect.disabled = isInfra;
+    if (isInfra) agentUnitSelect.value = '';
+  }
+  if (agentSelect) agentSelect.addEventListener('change', syncUnitSelectForSelection);
+
   let agentFetching = false;
   let agentLastFetch = 0;
   async function fetchAgent() {
@@ -86,7 +113,8 @@ import { createTabStrip } from '@hive/shared/tabs.js';
     agentFetching = true;
     agentOutput.textContent = 'fetching…';
     if (agentFetchTs) agentFetchTs.hidden = true;
-    const unit = agentUnitSelect ? agentUnitSelect.value : '';
+    const isInfra = infraContainerNames.has(name);
+    const unit = (!isInfra && agentUnitSelect) ? agentUnitSelect.value : '';
     const params = new URLSearchParams({ lines: '500' });
     if (unit) params.set('unit', unit);
     try {

hive-c0re/src/dashboard/journal.rs

@@ -1,10 +1,13 @@
 //! Journal-read endpoints for the dashboard.
 //!
-//! `GET /api/journal/{name}` reads a managed container's journal via the
-//! root helper (`journalctl -M`, delegated to hive-priv since hive-c0re is
-//! unprivileged). `GET /api/journal-host` reads host-side journald, both
-//! gated by an allow-list of known units so arbitrary unit names can't be
-//! probed. Operator-only by virtue of the dashboard binding host-only.
+//! `GET /api/journal/{name}` reads a managed agent container's journal, OR
+//! one of the four hive infra containers (`hive-ci`, `hive-forge`,
+//! `hive-gateway`, `hive-matrix` — [`hive_priv_sock::InfraContainer`] is the
+//! allowlist), via the root helper (`journalctl -M`, delegated to hive-priv
+//! since hive-c0re is unprivileged). `GET /api/journal-host` reads
+//! host-side journald, both gated by an allow-list of known units so
+//! arbitrary unit names can't be probed. Operator-only by virtue of the
+//! dashboard binding host-only.
 
 use axum::{
     extract::Path as AxumPath,
@@ -33,10 +36,23 @@ pub(super) struct JournalQuery {
 /// Operator-only by virtue of the dashboard being host-bound. hive-c0re
 /// runs unprivileged (privsep), so the `-M` read — which enters the
 /// container namespace and needs root — is delegated to hive-priv.
+///
+/// `name` is either a managed agent name (`iris`, optionally already
+/// carrying the `h-` prefix) or one of the four infra container names
+/// (`hive-ci` / `hive-forge` / `hive-gateway` / `hive-matrix` — see
+/// [`hive_priv_sock::InfraContainer`]). Infra containers don't run the
+/// per-agent hive daemons, so `unit` is ignored for them — always the
+/// full machine journal.
 pub(super) async fn get_journal(
     AxumPath(name): AxumPath<String>,
     axum::extract::Query(q): axum::extract::Query<JournalQuery>,
 ) -> Result<Response, ProblemDetails> {
+    let lines = q.lines.unwrap_or(500).min(5000);
+
+    if let Ok(infra) = name.parse::<hive_priv_sock::InfraContainer>() {
+        return read_journal_response(infra.unit_name(), None, lines).await;
+    }
+
     // Defense-in-depth format check so weird chars never reach the
     // shellout below — the `lifecycle::list()` existence check would
     // catch them anyway, but rejecting at the boundary keeps the
@@ -54,7 +70,6 @@ pub(super) async fn get_journal(
         return Err(ProblemDetails::from_status_code(StatusCode::NOT_FOUND)
             .with_detail(format!("journal: no managed container {prefixed:?}")));
     }
-    let lines = q.lines.unwrap_or(500).min(5000);
     let unit = match q.unit.as_deref().filter(|s| !s.is_empty()) {
         Some(u) => {
             // accept any of the per-container hive daemons [.service] —
@@ -78,8 +93,21 @@ pub(super) async fn get_journal(
         }
         None => None,
     };
+    read_journal_response(&prefixed, unit, lines).await
+}
+
+/// Shared `journalctl -M <machine> [-u <unit>]` shellout + response
+/// formatting for [`get_journal`], factored out so the infra-container
+/// branch (no `unit` filtering) and the agent-container branch (allow-listed
+/// `unit` filtering) don't duplicate the priv-client call + stdout/stderr
+/// combining.
+async fn read_journal_response(
+    machine: &str,
+    unit: Option<String>,
+    lines: u32,
+) -> Result<Response, ProblemDetails> {
     match crate::priv_client::read_container_journal(
-        &prefixed,
+        machine,
         hive_priv_sock::JournalQuery {
             lines,
             boot: true,

hive-c0re/src/forge/mod.rs

@@ -169,9 +169,15 @@ pub(crate) fn api(token: &str) -> Result<Forgejo> {
 ///
 /// Called by both `ensure_all()` (startup sweep) and `rebuild_agent`
 /// (per-rebuild) so the two paths stay equivalent.
-pub async fn sync_agent(name: &str, core_token: Option<&str>) {
+/// Returns `true` if all steps succeeded, `false` if any step failed. The
+/// caller can use the return value to aggregate per-agent failures into a
+/// dashboard warning (see [`ensure_all`]); the rebuild path ignores it and
+/// relies on the journal `warn!` lines alone (a rebuild is its own retry).
+pub async fn sync_agent(name: &str, core_token: Option<&str>) -> bool {
+    let mut ok = true;
     if let Err(e) = ensure_user_for(name).await {
         tracing::warn!(%name, error = ?e, "forge: ensure_user failed");
+        ok = false;
     }
     // Align email to match the git user.email set by meta::render_flake
     // so commits link to the agent's Forgejo profile. Best-effort;
@@ -188,9 +194,11 @@ pub async fn sync_agent(name: &str, core_token: Option<&str>) {
     // was down.
     if let Err(e) = ensure_config_repo(name).await {
         tracing::warn!(%name, error = ?e, "forge: ensure_config_repo failed");
+        ok = false;
     }
     if let Err(e) = push_config(name).await {
         tracing::warn!(%name, error = ?e, "forge: push_config failed");
+        ok = false;
     }
     // Grant read-only access to core/meta and wire the `meta` remote
     // into the proposed repo so agents can fetch their deployment context.
@@ -198,9 +206,11 @@ pub async fn sync_agent(name: &str, core_token: Option<&str>) {
         && let Err(e) = meta_read_access(name, token).await
     {
         tracing::warn!(%name, error = ?e, "forge: ensure_meta_read_access failed");
+        ok = false;
     }
     if let Err(e) = ensure_meta_remote(name).await {
         tracing::warn!(%name, error = ?e, "forge: ensure_meta_remote failed");
+        ok = false;
     }
     // Grant read-only access to internal/docs so the agent can clone
     // the operator-curated shared skills/runbook repo. Best-effort.
@@ -208,8 +218,10 @@ pub async fn sync_agent(name: &str, core_token: Option<&str>) {
         && let Err(e) = shared_docs_access(name, token).await
     {
         tracing::warn!(%name, error = ?e, "forge: shared_docs_access failed");
+        ok = false;
     }
     // internal/knowledge is public — no per-agent collaborator grant needed.
+    ok
 }
 
 /// The `core_token.is_some()` half of [`ensure_all`]: orgs, teams, the meta
@@ -346,11 +358,27 @@ pub async fn ensure_all() {
         );
         return;
     };
+    let mut sync_failed: Vec<String> = Vec::new();
     for c in containers {
         let Some(name) = c.strip_prefix(crate::lifecycle::AGENT_PREFIX) else {
             continue;
         };
-        sync_agent(name, core_token.as_deref()).await;
+        if !sync_agent(name, core_token.as_deref()).await {
+            sync_failed.push(name.to_owned());
+        }
+    }
+    if !sync_failed.is_empty() {
+        // Use static_kind to mint a `&'static str` key from the failed-agent
+        // list (bounded leak: one per hive-c0re boot, not per request).
+        let key = static_kind(format!("forge_sync_agent_{}", sync_failed.join("_")));
+        crate::warnings::set_boot_warning(
+            key,
+            "warn",
+            format!(
+                "forge: per-agent sync failed for: {} (see journal for per-step detail)",
+                sync_failed.join(", ")
+            ),
+        );
     }
 }