* refactor(matrix): adopt core claimable dedupe for inbound events
* fix(matrix): migrate legacy inbound dedupe markers via doctor
* fix(matrix): single-pool dedupe namespace and capacity-aware legacy import
* fix(matrix): keep unreadable legacy dedupe JSON in place during doctor import
* feat(dashboard): modular dashboard — workspace store, Workspaces tab, sandboxed custom widgets
Squashes openclaw/openclaw#101094 + #101097 + #101098 onto current main and
applies the maintainer review fixes to the backend control plane.
Co-authored-by: 100yenadmin <239388517+100yenadmin@users.noreply.github.com>
* fix(dashboard): UI review fixes — grid, error boundary, embed sandbox, locale
* fix(dashboard): make the CLI and agent broadcasts actually reachable
Three defects only a live run surfaces, all invisible to the unit suites:
- The plugin claimed the CLI command name `dashboard`, which core already owns
(it opens the Control UI). A plugin CLI group that overlaps a core command is
dropped at registration behind a `logger.debug`, so the entire CLI face was
unreachable while `cli.test.ts` kept passing against its own Commander
program. Renamed to `openclaw workspaces`, matching the tab it drives.
- The manifest never declared `activation.onCommands`, so the CLI root resolved
to no owning plugin even once the name was free.
- `dashboard.widget.approve` needs `operator.approvals`; the CLI asked for
`operator.write` on every call. It now requests the approvals scope only for
the approve call, matching `operator-approvals-client.ts`.
Also: agent tools resolved their broadcast from the plugin runtime's
gateway-request scope, an AsyncLocalStorage set only around gateway RPCs and
plugin HTTP routes. An agent turn started from a channel, cron, or heartbeat
therefore wrote the document without emitting `plugin.dashboard.changed`, so an
open Control UI never saw the edit — the feature's headline promise. The gateway
broadcast is server-lifetime, so the plugin now remembers it in a single slot and
agent tools fall back to it.
* docs(web): document dashboard workspaces, provenance, and the custom-widget sandbox
* fix(dashboard): agent-tool ergonomics + close two approval-boundary gaps
From a source-blind agent driving the dashboard_* tools with nothing but their
schemas, and from a Codex review of the hardening delta.
- dashboard_widget_update could never succeed. It passed its whole parameter
record to the patch reader, whose allowlist rejects the very `tab`/`id` keys
the tool's own schema marks required, so every call died on
"unexpected param: tab". Its test only ran Value.Check against the schema and
never executed the tool.
- dashboard_data_read surfaced an `rpc` binding as a thrown error, though its
description promised `binding_client_resolved`. It now returns that as a
result the model can act on.
- Valid widget kinds and the rpc allowlist were undiscoverable: a model saw only
"builtin:<name> or custom:<name>" and "Allowlisted gateway read method", then
brute-forced ~40 calls against errors that named no alternatives. Both schemas
and both validator errors now enumerate them, and the kind description says
what each builtin renders and which binding id it reads. widget_move documents
that grid and toTab are exclusive; widget_scaffold says an operator must
approve, because no agent tool can.
- workspace.replace could mint a pending registry entry for a name that was
never scaffolded. An operator could then approve a widget whose code did not
exist yet, and the agent could write it afterwards. Registry entries now come
from dashboard_widget_scaffold and nowhere else, and approve refuses a name
with no manifest on disk.
- dashboard.widget.approve answered with the whole workspace document, so a
connection holding only operator.approvals could read it through the approvals
door. It now returns the registry entry it changed.
* fix(dashboard): approval pins the code it approves
Codex review found the scaffold-before-approval gate still nameable rather than
binding: approve only proved that widget.json parsed, and the Control UI loaded a
hardcoded index.html rather than the manifest's entrypoint. An agent could
scaffold a widget, win approval on an innocuous or absent entrypoint, then write
the real payload afterwards — code appearing after the human said yes.
Approval now hashes every servable file in the widget directory and stores the
digests on the registry entry, refusing a manifest whose declared entrypoint is
missing. The asset route re-hashes each file it reads and 404s anything that does
not match, so a file edited or added after approval never reaches a browser. The
Control UI loads the manifest's entrypoint, which is the file that was hashed.
The content-type allowlist moves to manifest.ts so the set of files approval
hashes and the set the route can serve cannot drift apart.
Proof, against a running gateway: scaffold -> 404, approve -> 200, rewrite
index.html -> 404, add late.js -> 404.
* fix(dashboard): parse the approved manifest from the bytes that were hashed
Codex found a TOCTOU in the approval path: it loaded and validated widget.json,
then walked the directory again to compute the digests. An agent could swap
widget.json between the two reads, so the operator validated one entrypoint while
the digest froze — and the Control UI later mounted — a different one.
snapshotApprovedWidget now reads the widget directory once: it hashes every
servable file, parses the manifest out of the same widget.json bytes it hashed,
and requires the declared entrypoint to be among them.
Proof, against a running gateway: approve -> index.html 200; rewrite widget.json
to point at evil.html and drop evil.html in -> both 404.
* fix(dashboard): cap approval asset reads; bound the grid fallback search
Two findings from the fourth Codex pass.
Approval hashes agent-authored files that are untrusted until it runs, and read
each one into memory with no size check — dropping one huge .png into a scaffold
directory would stall or OOM the gateway during approve. Sizes are now checked
before the read, with a 2 MB per-file and 8 MB total cap.
nearestFreeSlot searched one band below the lowest occupied row, so a crowded
layout near the bottom could return y=500 as the closest free slot: a placement
the store rejects, which the UI applies optimistically and then snaps back. The
search now stops at the last row a widget of that height can legally occupy.
* fix(dashboard): refuse oversized widget assets before reading them
Approved widget files stay writable and the asset route is unauthenticated, so
swapping an approved small file for a very large one made every GET buffer the
whole file before the digest check rejected it. The route now refuses anything
past the same per-file cap approval enforces, on the stat it already performs.
* fix(dashboard): enforce widget approval boundaries
* docs(changelog): note modular dashboard workspaces
* fix(dashboard): enforce static custom-widget data boundary
* fix(dashboard): satisfy UI lint
* test(dashboard): avoid legacy proto access
* feat(dashboard): make plugin opt-in
* docs(dashboard): refresh workspaces map
* refactor(workspaces): standardize plugin naming
* fix(workspaces): make widget prompt sends idempotent
* docs(workspaces): fix internal path references
* test(workspaces): make prompt assertion lint-safe
* test(workspaces): type prompt request mock
* fix(workspaces): harden approval and binding boundaries
* test(workspaces): complete stale binding client mock
* fix(workspaces): harden widget file boundaries
* fix(workspaces): scope custom widget capabilities
* fix(workspaces): align approval provenance
* fix(workspaces): close branch contract gaps
* test(workspaces): complete builtin context fixtures
* fix(workspaces): aggregate overview usage
* chore(workspaces): defer release note
* chore(workspaces): refresh i18n metadata
---------
Co-authored-by: 100yenadmin <239388517+100yenadmin@users.noreply.github.com>
* fix(discord): guard progress draft collapse to prevent re-collapse on cleaned stream
Commit 86ea382 added a progressDraftStartedBeforeFinal latch that snapshots
hasStarted at markFinalReplyStarted time. This correctly preserves collapse
eligibility when an abort cancels the compositor gate before
shouldCollapseProgressDraft runs.
However the latch was never cleared. After markPreviewFinalized turned the
draft into a summary, hasProgressDraftStarted remained true permanently. In
multi-tool turns with multiple final-payload deliveries, this caused
shouldCollapseProgressDraft to re-trigger collapse on an already-cleaned or
sealed draft stream, corrupting downstream delivery state and causing tool
results to render as image content blocks instead of text.
Fix: add a progressDraftCollapsed guard set by markPreviewFinalized that
prevents hasProgressDraftStarted from returning true after the draft has been
collapsed into a summary. Reset the guard on handleAssistantMessageBoundary
so followup/queued turns start fresh.
Per Discord channel session, all tool results rendered as image content
blocks. Model received (see attached image) placeholders instead of text.
Fixes#100782
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(discord): use valid table/chunk mode literals in collapse-guard tests
* test(discord): add abort-race gate-cancellation latch survival test
* fix(discord): consume progress draft collapse once
Co-authored-by: xialonglee <li.xialong@xydigit.com>
* fix(discord): retain progress receipt on retry
Co-authored-by: xialonglee <li.xialong@xydigit.com>
* fix(discord): rearm queued progress drafts
Co-authored-by: xialonglee <li.xialong@xydigit.com>
* test(discord): assert progress lifecycle outcomes
Co-authored-by: xialonglee <li.xialong@xydigit.com>
* fix(discord): reset queued progress turn state
Co-authored-by: xialonglee <li.xialong@xydigit.com>
* test(discord): model recreated progress drafts
Co-authored-by: xialonglee <li.xialong@xydigit.com>
* fix(discord): serialize progress draft rotation
Co-authored-by: xialonglee <li.xialong@xydigit.com>
* fix(discord): isolate queued draft generations
Co-authored-by: Peter Lee <xialonglee@users.noreply.github.com>
* test(discord): align synchronous draft rotation
Co-authored-by: Peter Lee <xialonglee@users.noreply.github.com>
* fix(discord): distinguish draft rotation ownership
Co-authored-by: Peter Lee <xialonglee@users.noreply.github.com>
* chore: keep release notes in pull request
---------
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* fix(discord): stop retrying non-idempotent sends on post-connect-ambiguous errors
Discord's outbound retry runner treated every transient transport error as
retryable, including ECONNRESET, ETIMEDOUT, AbortError, and the undici
headers/body/socket timeouts. Those errors can fire after Discord has already
received and created the message but before the client reads the response, so
replaying the non-idempotent createChannelMessage POST delivers the same
message twice while reporting success.
Add a pre-connect-only classifier (isRetryableDiscordPreConnectError) covering
errors that provably never reached Discord (DNS/connect refused/connect
timeout) plus rate-limit rejections, mirroring the Telegram send-path
carve-out. The retry runner now accepts a per-call nonIdempotent option; the
message-create and thread-create call sites opt in so ambiguous post-connect
errors surface to the caller instead of double-sending. Idempotent REST calls
(reactions, edits, DM-channel lookup, uploads) keep the broader transient set.
* fix(discord): retry nonce-enforced message creates
* fix(discord): scope create retries by endpoint
* fix(discord): stabilize sticker and poll create nonce across retries
* fix(discord): tighten create retry safety
* fix(discord): preserve nonce-safe retries
* fix(discord): scope nonce fields to message creates
* test(discord): cover nonce-safe direct sends
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* fix(google-meet): force English UI when reusing existing Chrome tab
* fix(google-meet): narrow reused tab before forcing English UI
* test(google-meet): assert /navigate target id is adopted in reused tabs
* fix(google-meet): preserve active call tabs and normalize English before recovery
* fix(google-meet): preserve localized active tabs
* test(google-meet): avoid shadowing path import
* test(google-meet): align mocks with English-pinned tab reuse
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* fix(logs): clean up gateway and channel startup/shutdown log output
Scope Discord slash-command deploy REST diagnostics to command routes so
concurrent startup traffic (voice-state probes, channel lookups) keeps its
owner's error handling; make per-request deploy error lines verbose-only and
drop JSON bodies that only repeat message+code. Log allowlist summaries one
line per call so unresolved lines keep their timestamp/subsystem prefix, and
skip identity lookups that resolved to themselves. Remove embedded subsystem
prefixes, demote routine signal/shutdown/force/postbuild/diag chatter to
debug or verbose, merge the duplicate Control UI build notices, drop doctor's
duplicate backup line, and name the config surface or platform limit in the
transcripts autoStart and command-limit warnings.
Fixes#104163
* fix(slack): keep bare-name allowlist resolutions in startup log summary
Only omit identity lookups where the input already is the resolved id;
name-based lookups that translated to an id stay logged even when the
display name matches the input.
* fix(telegram): keep isolated-ingress readiness marker on the runtime log
test/e2e/qa-lab telegram-bot-token-runtime waits for this line via the
injected RuntimeEnv.log; verbose-only logging would deterministically time
out that live proof. Comment the contract at the call site.
