claude emits a started tick then a completed tick per plugin install;
previously each got its own row (loading... then done) instead of one
line updating in place. Same coalescing pattern already used for
thinking_tokens and the generic status tick.
Claude's generic `system/status` subtype carries no detail beyond the
bare label, and previously each tick got its own note row. During a
compaction pass (which emits a burst of these with no other signal
before the completed compact_boundary) this looked like a wall of
identical noise followed by silence, making a routine compaction look
stuck.
Collapse consecutive status ticks into one updating row, same pattern
already used for the thinking_tokens counter.
Both mdNode implementations (agent UI app.js, dashboard common.js)
assigned marked.parse() output straight to innerHTML with no
sanitizer. marked v5+ dropped its built-in sanitize option, and there
was no DOMPurify anywhere in frontend/, so markdown containing raw
HTML/script tags rendered live in the browser.
Both sinks receive untrusted input in practice: the agent UI's mdNode
renders recv tool_result bodies, assistant prose, and send/ask/answer
payloads sourced from peer agents and matrix-relayed messages (the
documented prompt-injection adversary); the dashboard's mdNode renders
agent-authored .md files served verbatim by GET /api/state-file
(the endpoint validates path, not content). Since the per-agent UI and
dashboard are same-origin behind the gateway with operator-authority
endpoints (approve/spawn/rebuild/destroy/answer-question), injected
script would run with the operator's session.
Fix: DOMPurify.sanitize() the marked.parse() output at both sinks
before assigning to innerHTML. Added dompurify as a dependency to
both the agent and dashboard npm workspaces, recomputed npmDepsHash
in nix/frontend.nix for the updated lockfile. Also corrected
docs/web-ui/shape.md, which claimed the markdown-rendering path was
XSS-safe by construction the same way the text-node-based linkify
path is — it isn't; it's safe because it's sanitized.
CSP hardening for the dashboard (no unsafe-inline) is a separate,
larger backend change (response headers in hive-c0re) and is left as
a fast-follow rather than folded into this fix.
- deploy-window gate (meta::exclusive) + path-limited meta commits:
a perm/lock/topology commit can no longer sweep an ApprovalDeploy's
staged flake.lock and neuter abort_deploy (regression test included)
- cancel surfaces now buffer terminal roll-ups the scheduler drains,
so a queued approval DAG cancelled by the operator resolves its
approval instead of dangling, and cancelled power ops revert their
wanted flip to the observed state
- hivectl restart / restart-all ride the queue (lease serialization,
transient guard) and restart sets wanted=Up like the old kill+start
- exactly one Rebuilt event per rebuild DAG, emitted at terminal
- StopForUpdate pre-seeds a missing agent_power row from the pre-stop
observation so a rebuild can't strand an unknown agent offline
- history trim keeps terminal fan-out parents with live children
- audit_log back on db::open; swarm.js badge for reconcile DAGs
each queue card now shows its DAG's node chain (per-node state, step,
build-log link), fixing 'queue jumps don't show on the dashboard'.
live-log panel keys off the running node's log. new
services.hyperhive.c0re.buildSlots option (default 1) threads the
concurrent nix-build count into serve.json.
xhigh and max are distinct reasoning-effort levels (see EFFORT_LEVELS
in events.rs); labelling xhigh as '(max)' implied they were the same
level, per issue #2258.
Empty arrays are truthy in JS so 'if (input.targets_add)' would emit
'+0 tgt' for an explicit []. Add .length guard so zero-element lists
are silently skipped — consistent with fmtArgsGeneric's [N] handling.
The MCP schema defaults both fields to null, making this theoretical,
but the guard is cleaner.
Scheduling tools and the two remaining request_* tools fell through to
fmtArgsGeneric. They already had appropriate icons (⏱️ / 📦) but the
arg display was multi-field verbose.
New fmtToolUse cases:
- request_init_config: 'request_init_config* iris'
- request_update_meta_inputs: 'request_update_meta_inputs* [nixpkgs, …]'
(or 'all' when inputs list is empty)
- list_schedules: 'list_schedules*()' (no args, explicit)
- cancel_schedule: 'cancel_schedule* #42 all' or '#42 [iris, dmatrix]'
- fire_schedule_now: 'fire_schedule_now* #42'
- edit_schedule: 'edit_schedule* #42 · body · interval · +2 tgt'
(lists which fields are being changed, not their values)
- request_schedule_prompt: 'request_schedule_prompt* → iris at 14:00Z +3600s'
Closes#2201
join_room and open_dm shared a fall-through case with
(fmtRoom(input.room) || fmtUser(input.user_id) || '?'). fmtRoom()
returns the string '?' when input.room is undefined — truthy — so
the fmtUser() fallback was never reached and open_dm always displayed
'open_dm* ?'.
Split into two separate cases: join_room reads input.room, open_dm
reads input.user_id. Both degrade to '?' via fmtRoom/fmtUser's own
null guard.
`mcp__hyperhive__remind` calls previously rendered as the generic
`fmtArgsGeneric` output — with a multi-field input that showed
`message: "..." · delay_seconds: 300`, burying the message after
a verbose field name.
New format: `remind* +5m "check on PR..."` (or `at HH:MMZ` for
absolute timestamps). The timing renders first so it's scannable
at a glance, followed by the first 60 chars of the message body.
Works for all three input shapes: delay_seconds, at_unix_timestamp,
and file_path-only (shows the path as the preview).
Several MCP tools appeared with the generic wrench icon (🔧) making
them hard to identify in the scrollback. fmtToolUse also lacked
specific formatters for some frequently-used tools, falling through
to fmtArgsGeneric.
Icons added:
- ack_until → ✅ (mark-as-read semantics)
- get_logs → 📜 (log viewer)
- get_host_journal → 📜 (journal reader)
fmtToolUse cases added:
- ack_until → "ack_until* ≤N" (message-id bound)
- get_logs → "get_logs* <agent> [NL]"
- get_host_journal → "get_host_journal* <container|unit> [/grep/] [NL]"
- restart/start/update → "restart* <name>" etc. (previously used
fmtArgsGeneric; now match kill's pattern)
Closes#2188.
compact_boundary events currently render as the generic "⚙ compact_boundary"
muted note. The event carries useful metadata — pre/post token counts,
duration and trigger — that are invisible to the operator.
With this change the row reads:
· ⚙ compact · manual · 772k→6k tokens · 101s
Fields rendered (all guarded — missing fields are silently omitted):
- trigger ("manual" or "auto")
- pre_tokens→post_tokens (formatted with k/M suffixes)
- duration_ms (ms or s)
Closes#2187.
Currently both system subtypes fall through to the generic muted
note renderer (⚙ <subtype>), giving the operator no insight into
what is happening.
plugin_install:
Render status explicitly — "loading…" on started, "✓ done" on
completed — so a slow plugin boot is visible in the scrollback
instead of two identical cryptic rows.
commands_changed:
Render the count of available slash commands in the summary row
and expand to the full list in a collapsible details block. The
list is most useful right after a fresh plugin_install so the
operator can see exactly which /commands are now on offer.
Both handlers sit immediately before the generic catch-all in the
system-subtype dispatch in renderStream, preserving the existing
fall-through for other subtypes (context_window_exceeded etc.).
Closes#2183.
Two operator-visible strings and one code comment still referenced 'the
manager' in the approval card UI:
- init_config card description: 'manager customises agent.nix before
spawn' -> 'submitting agent customises agent.nix before spawn'
- deny form data-prompt: 'sent to manager' -> 'sent to submitter'
(the deny note goes to the ApprovalResolved submitter, not a fixed
manager role)
- code comment: 'surfaced to the manager via' -> 'surfaced to the
submitting agent via'
- builds.html meta text: 'manager learns each outcome' -> 'the
submitting agent learns each outcome'
The 'manager' role is no longer structural — root-ness is topological.
Any agent with the approvals tool group can submit; the note/event routes
to the submitter.
The QueueKind enum has graceful_stop, start, and stop variants but the
dashboard QUEUE_KIND_GLYPH map only covered rebuild/meta_update/spawn/
destroy/restart/startup_sweep/perm_change — the three missing kinds
rendered as '?' in the build queue UI.
This became visible with the deferred start-after-rebuild change: a fast-lane
Start entry now appears as a child of its Rebuild parent, but showed the
fallback '?' glyph.
Glyphs assigned: graceful_stop=⏹, start=▶, stop=■.
Also update the dashboard.md kind-glyph list to include all nine kinds and
note that deferred start-after-rebuild entries also use parent_id grouping.
fix(dashboard): update 'view queue' link to /builds.html
The build queue moved to its own /builds.html page. The queue-summary
'view queue →' link in tabs.js still pointed at /core.html, so clicking
it landed on the wrong page.
fix(broker): filter agent inbox to unread (acked_at IS NULL)
recent_for was returning all messages regardless of ack state, so the
agent inbox showed everything even after 'mark all read'. Now filters
to acked_at IS NULL — mirroring exactly what mark_all_read drains —
so the inbox empties on reload after the operator drains it.
When a tool call fails, claude wraps the result text in
<tool_use_error>...</tool_use_error> XML tags. The terminal was
displaying these raw, producing output like:
'<tool_use_error>File has not been read yet.</tool_use_error>'
Fix renderToolResult in app.js:
- Check c.is_error on the tool_result content block.
- Strip the <tool_use_error>...</tool_use_error> wrapper from the text.
- Render error results with a '✗' prefix under '.tool-result.error'
(flat, ≤120c) or '.tool-result-block.error' (<details>, longer text).
Add .live .tool-result.error { color: var(--red); } to terminal.css
so error results are visually distinct (red, same as turn-end-fail).
Update terminal-rendering.md row taxonomy to document the two new
error row classes.
Closes#2104.
Remove or shorten comment blocks that:
- explain where code used to live (impl history)
- duplicate rationale already in docs/web-ui.md
- contain speculative/future-work notes
Changes:
- tabs.js: drop "SYST3M tab used to render this" history; keep badge purpose
- schedules.js: drop "used to be a card layout" history + speculative nit note
- stream-worker.js: replace 10-line problem-statement with docs pointer
(rationale already in docs/web-ui.md SSE multiplexing section); trim
future-work parenthetical from subscription-tracking comment
- system-sections.css: drop extraction-history prose; keep what it does
- core.css: drop move-history prose; keep what it imports and why
Part of issue cleanup per operator feedback.
New standalone page /builds.html consolidating rebuild queue, meta
inputs, and build log history into one place, with a new 'Builds' home
tile linking to it. Addresses mara's request: new sub-page with a new
home tile, all three items moved there.
Changes:
- builds.html: new page with three sub-tabs: R3BU1LD QU3U3, M3T4 1NPUTS,
BUILD L0GS. Same minimal-chrome header + createTabStrip pattern as core
and logs pages.
- builds.js: new bundle combining rebuild queue renderer (from core.js),
meta inputs renderer (from core.js), rebuild-live-log renderer (from
core.js), and build log history renderer (from logs.js). Deep-links to
/builds.html?id=N#buildlogs. Count pill id: builds-tab-count-rebuild.
BUILD L0GS tab lazy-loads on first activation.
- builds.css: @imports system-sections.css (rebuild queue + meta inputs +
live-log styles) and logs.css (build-logs-* component styles).
- build.mjs: register builds.js, builds.css, builds.html.
- core.html: remove R3BU1LD QU3U3 + M3T4 1NPUTS tabs (now on builds.html).
Default tab changes to K3PT ST4T3.
- core.js: remove renderMetaInputs, renderRebuildQueue + helpers,
renderRebuildLiveLog + live-log state, elapsed-time tickers,
updateRebuildCount, and the rebuild_queue/meta_inputs SSE handlers.
Remove openBuildLogStream + util imports no longer needed.
- core.css: remove .rebuild-live-log-* rules (moved to system-sections.css
so builds.css can share them via @import).
- system-sections.css: add .rebuild-live-log-* styles (moved from core.css);
update comment to mention builds.html.
- logs.html: remove BUILD tab + pane (moved to builds.html).
- logs.js: remove fetchBuild(), fmtTs, fmtDuration, openBuildLogStream
import, and rebuild_queue_changed SSE debounce. Default tab: 'agent'.
SSE stream retained for audit_entry_added live-appends.
- index.html: add Builds tile (🔨, rebuild queue · meta inputs · build
logs); update Core tile desc to 'kept state · container load'; update
Logs tile desc to remove 'build'.
Add GET /api/permissions/stale endpoint that returns agent names with
explicit capability/tool-group JSON entries but no live container AND
no kept-state tombstone. Ghost detection is now entirely server-side —
one authoritative call, no client-side roster cache, no staleness window.
The previous client-side approach in core.js made three parallel API
calls (GET /api/capabilities, GET /api/tool-groups, GET /api/state) and
filtered the result against a module-level `liveContainerNames` Set
populated only on cold load and form submits. Any container lifecycle
event (spawn, destroy) while core.html was open left `liveContainerNames`
stale, risking a false-positive ghost entry for a live container.
Changes:
- permissions.rs: add `get_stale_permissions` handler + `StalePermsResponse`
struct. Computes live roster (containers_snapshot), tombstone set
(Coordinator::kept_state_names), explicit perm names (capabilities::read
+ tool_groups::read), then returns the difference sorted.
- dashboard.rs: register GET /api/permissions/stale.
- core.js: replace the three-call client-side logic in
`fetchAndRenderStalePerms` with a single fetch to /api/permissions/stale.
Remove `liveContainerNames` state + its syncFromSnapshot population.
Adds a "stale permission entries" sub-section within the K3PT ST4T3 pane
on /core.html showing agents with explicit capability/tool-group JSON
entries but no live container — typically renamed or manually-deleted agents
whose entries persisted (e.g. the old "root" manager name after rename to
"ruth").
Each ghost agent gets a "✕ clear perms" button that calls
DELETE /api/permissions/{agent} (added in the prior commit). Lazy-loaded
on first K3PT ST4T3 tab activation; auto-refreshes on capabilities_changed
and tool_groups_changed SSE events.
core.js: track liveContainerNames from /api/state.containers; add
renderStalePerms + fetchAndRenderStalePerms; hook tab onShow + SSE handlers.
system-sections.css: new .tombstones-stale-* selectors for the ghost list
rows and error message.
core.html: add #tombstones-stale-perms div inside the K3PT ST4T3 pane;
expand comment to describe both sub-sections.
Read the persisted model name from each agent's harness state file
(harness/hyperhive-model) and surface it as a small blue badge on
the container row in the SW4RM tab.
- container_view.rs: add `active_model: Option<String>` to
ContainerView; populated by new `read_active_model` helper that
reads harness/hyperhive-model; only set when container is running
(stale model info from a stopped agent is misleading)
- container_view.rs: add active_model to ContainerView literal in
host_stats test helper
- tabs.js: render badge-model chip after needs-update, before
reminders; add active_model to the row fingerprint so re-renders
fire on model change
- common.css: add .badge-model (blue, 80% opacity — informational)
Adds pause/resume support for scheduled prompts.
Backend:
- New paused_at_unix column on scheduled_prompts table (added via
ALTER TABLE migration so existing databases are upgraded on first
start). The due-rows index is dropped and recreated to also exclude
paused rows so the worker never fires them while paused.
- Worker's due() query gains AND paused_at_unix IS NULL filter.
- New pause(id) and resume(id) methods on ScheduledPrompts; both are
idempotent and refuse cancelled rows.
- New POST /api/schedules/{id}/pause and /api/schedules/{id}/resume
dashboard endpoints (operator-direct, no approval gate). Both emit
a schedules snapshot on success so the tab updates live.
- WireSchedule gains paused_at_unix: Option<i64> so the frontend can
render the state without an extra fetch.
Frontend:
- Paused rows render with a distinct row class + muted opacity.
- The next-fire cell shows a yellow pause glyph + tooltip with the
paused-since timestamp and the would-have-fired time.
- Actions column: pause/resume toggle button (⏸/▶) beside fire/edit/cancel.
Fire-now is disabled while paused (resume first).
- Sort order: active → paused → cancelled (paused slot keeps schedules
visible without mixing them into the active top section).
- pauseSchedule() / resumeSchedule() async functions POST to the new
endpoints and refresh the table on success.
Both remove_agent() calls now run unconditionally for maximum partial
cleanup, but any I/O error is returned as HTTP 500 instead of silently
200-ing — so the frontend's !resp.ok path fires and the operator sees a
meaningful error rather than the stale row reappearing unchanged.
Also add a clarifying comment on isStale in permissions.js explaining
that containersState is keyed from nixos-container list (which includes
stopped-but-configured containers), so a temporarily-stopped agent is
not treated as stale — only destroyed/renamed agents are absent.
The P3RM1SS10NS tab showed agents that no longer exist in the live
container roster — e.g. an agent named 'root' that was renamed or
destroyed but still had explicit entries in tool-groups.json and/or
capabilities.json. The roster-union behaviour is intentional for
temporarily-stopped agents, but stale entries from renamed/destroyed
agents are confusing.
Backend (dashboard/permissions.rs):
- New DELETE /api/permissions/{agent} handler that bypasses the live-
roster guard (intentionally — that's the point). Calls
tool_groups::remove_agent + capabilities::remove_agent to clear both
JSON files, then emits live SSE snapshots so the tab updates without
a page reload. Format-checks the agent name but does not require it to
be in the containers snapshot.
Frontend (permissions.js):
- renderCapabilities / renderToolGroups now cross-reference agentNames
against containersState (the live roster, already imported). Agents
not in the live roster get an isStale flag.
- Stale rows get a '(not running)' label and a '✕ remove' button that
calls clearStaleAgent() — a new async helper that DELETEs the stale
entry and re-fetches both perm tables.
- Non-stale agents without explicit assignments still get '(default)'.
CSS (dashboard.css):
- .perm-row-stale (reduced opacity), .perm-stale-label (muted small
text), .perm-remove-btn (small red-bordered button) + disabled state.
The CSS for the rebuild-queue live-log panel targets .rebuild-live-log
(border, border-radius, margin-top, background) but the HTML element only
had id="rebuild-live-log" — no class. As a result the panel box styles
never applied and the live log rendered unstyled (no border, no background,
no visual separation from the queue rows).
Fix: add class="rebuild-live-log" to the element so the CSS selector
matches.
Both kinds fell through to the spawn branch in renderApprovals, showing
a misleading 'spawn' chip and agent-spawn body text. Mara saw a meta-input
bump render as a spawn card for agent damocles and denied it.
Backend (dashboard.rs):
- Add commit_ref: None to the MergeConfigPr arm (struct was incomplete).
All arms of ApprovalView now initialise every field.
Frontend (call.js):
- Add isUpdateMeta / isSchedule booleans alongside the existing kind flags.
- Glyph: update_meta_inputs gets ↻, schedule_prompt gets ⏱.
- Kind chip: 'meta-update' / 'schedule' (no kind-spawn class for either).
- Body: update_meta_inputs parses commit_ref as JSON Vec<String> and shows
'bump flake inputs: foo, bar' or 'bump all flake inputs'; schedule_prompt
parses SchedulePromptPayload and shows targets + first-fire time + cadence
+ a truncated body excerpt.
- History row: add 'meta-update' and 'schedule' cases (were both 'spawn').
- Import fmtDuration from util.js (needed for schedule cadence display).
Completes the merge_config_pr approval-card link for live-added approvals.
The ApprovalAdded SSE event now carries pr_number (only for
merge_config_pr); applyApprovalAdded was dropping it, so a live-added
merge_config_pr card showed the sha but not the 'review PR on forge' link
until a cold /api/state reload. Carry pr_number through the same way as
sha_short.
Follow-up to the destroy-guard removal — the bootstrap/root container is
now destroyable end-to-end. The frontend already offered DESTR0Y/PURG3
for every container; the only manager-gating left was a stale doc-comment
('destroy/purge hidden for the manager') and a vestigial bare block
wrapping the menu appends. Drop both (no behaviour change), and update the
dashboard.md menu doc: 'disabled for the bootstrap container' -> available
for any container (hive-c0re recreates it on next startup if destroyed).
renderApprovals had no merge_config_pr case, so those approvals
mis-rendered as generic spawn cards (spawn chip, no sha, 'container will
be created' body). Add a dedicated branch:
- distinct glyph + 'merge-pr' kind chip
- show the reviewed PR-head sha (sha_short), like apply_commit
- a 'review PR on forge' link built from pr_number (now on ApprovalView),
gated on forge_present, mirroring the apply_commit 'commit on forge' link
- no config-diff side-panel (apply_commit-only for now)
History rows get the 'merge-pr' chip too.
Adds a Window::All option to the per-agent stats page and the hive-wide
rollup, selectable as a new 'all' tab on both.
- hive-ag3nt (per-agent, time-bucketed): All ranges from MIN(started_at)
to now (fallback to now on an empty table) with an adaptive bucket
width laddered by span — hourly <=2d, daily <=90d, weekly <=2y, 30-day
beyond — so the trend series stays bounded (~<=104 buckets) at any age.
- hive-c0re (swarm rollup, not time-bucketed): All sets from=0 so the
aggregate covers every recorded turn across all agents.
- frontend: an 'all' button on both the agent stats and dashboard hive
stats window selectors (createTabStrip + fetch already pass the window
string through, so no JS change needed).
cargo check passes on both crates; FE builds clean. Fixes#1919.
