Commit Graph
10 Commits
Author SHA1 Message Date
kb 8a349c6045 fix(muse): keep subagent runs alive, finished, and on the configured model
Muse projects a subagent's approvals onto the parent with the child's view
cursor. The plugin stored that cursor as the parent's recovery anchor, so the
two-minute silence repair paged the parent from a cursor Muse rejects
("unknown cursor anchor") and the failed repair failed the whole session.
Projected events no longer move the parent's cursor, and a failed silence
repair now logs and retries at the next silence.

Muse sends no turn events for a subagent and leaves its parent item
inProgress after it finishes, reporting the end in controlStatus. Finished
subagents therefore stayed "running" in the Subagents track and on their
timeline card until the parent's session closed. The plugin now ends the
child's turn, and settles the card, when the item reports resultReady or
closed, or a terminal status.

session/resume takes no model or approval mode, and Muse 1.4.7 reports its
default model after a resume while it keeps running the configured one. The
plugin adopted and persisted the reported model. After a resume it now
re-applies the configured model and approval mode when Muse reports others.
2026-11-03 12:03:17 +00:00
Mohamed Boudra 3e2d77723e fix(usage): scope Muse and agent logins, show every failed host login (#6156)
* fix(muse): show usage only for the agent account

Usage discovery ignored session scope and exposed all remembered accounts to unrelated agents. Resolve session accounts from their own launch environment while preserving global discovery.

* test(muse): own the usage scope regression in the plugin suite

* fix(usage): scope agent reports to their own logins

Account grouping also merged fetch inputs and cache entries across agent and host queries, hiding the agent login failure behind another harness. Preserve scoped login sets, collect every host failure with source-owned harness labels, and keep refresh results within their requesting view.

* test(usage): expect harness labels in discovered accounts
2026-10-05 18:41:54 +01:00
Mohamed Boudra bc6ddbfa97 fix(muse): restore earlier messages after daemon restart (#6159)
MSP view cursors request only a suffix. A fresh provider session rebuilds an empty timeline, so it must resume without a cursor to recover earlier messages. Keep gap recovery cursors in memory and omit them from durable session state.
2026-10-05 17:47:04 +01:00
Mohamed Boudra 9760292650 fix(providers): launch plugin CLIs on Windows and tolerate Muse startup (#6151)
* fix(providers): launch plugin CLIs on Windows and tolerate Muse startup

Plugin providers bypassed the core Windows launcher and the plugin host exposed an empty server SDK module. Share the launcher through the SDK and bind it in the host. Give Muse initialization its own deadline because native startup traverses session history before answering MSP.

* fix(providers): terminate Windows launcher process trees

A command-script launcher leaves the CLI behind cmd.exe. Expose forced process-tree termination in the server SDK, use it for Muse and SDK ACP teardown, and bound Muse closing so inherited streams cannot leave callers waiting forever.

* test(muse): keep platform setup outside the teardown assertion

* refactor(plugin): expose process capabilities without quoting helpers

* fix(providers): close process trees on uncorrelated protocol errors
2026-10-05 17:10:02 +01:00
paseo-bot[bot]andMohamed Boudra d826d8109f fix(muse): load models whose effort tiers carry descriptions (#6021)
* fix(muse): load models whose effort tiers carry descriptions

Muse 1.4.2 sends reasoningEffortVariants as described tiers
({ tier, description? }) and the complete effort list on variants, as
the bundled MSP types declare. The catalog schema had the two shapes
swapped, so any model with a described tier failed the whole model/list
parse: the model picker stayed empty and agent creation failed.

Parse both fields as declared, offer the efforts from variants, and
show Muse's tier descriptions on the matching options.

* fix(muse): leave effort descriptions out until the pickers show them

* fix(muse): keep unused catalog metadata outside validation

Described effort tiers are presentation metadata, not selectable efforts. Validating their shape blocks both discovery and session startup even though Paseo never consumes them. Normalize the complete variants list at the wire boundary and cover session readiness and selecting max.

* fix(muse): align consumed session and event wire fields

---------

Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Mohamed Boudra <boudra.moha@gmail.com>
2026-10-05 16:12:12 +01:00
Mohamed Boudra 73d25317a4 fix(usage): derive window names and unify pin selection (#5903)
* fix(usage): derive window names and unify pin selection

Quota slots do not identify a period. Derive numeric windows and IDs from reported duration and quota scope; keep unknown lengths neutral. Resolve source defaults and explicit pins through one model, with a persisted empty selection distinct from defaults.

* fix(plugins): declare the duration helper host requirement

* fix(usage): verify persisted selections at the storage boundary
2026-10-02 11:18:13 +01:00
Mohamed Boudra 05e94ab52e Keep usage logins visible and explain unavailable quota (#5876)
* fix(usage): keep existing logins visible and explain unavailable quota

Credential discovery must preserve accounts whose login exists even when it has expired. Keep vendor failures on the discovered account and isolate discovery failures without inventing provider cards.

* test(usage): align credential route tests with account discovery
2026-10-02 00:30:11 +01:00
Mohamed Boudra e939dd4c7a Streamline sidebar usage and stacked sheet interactions (#5805)
* feat(app): streamline sidebar usage and stacked sheets

* fix(app): restore compact sidebar navigation clicks

The close button container emitted box-none through Unistyles as invalid CSS and intercepted New workspace clicks. Let React Native apply click-through behavior at the view boundary. Keep the import journey on its existing New workspace screen and assert only the visible transcript.

* fix(app): cap menu sheets at eighty percent height

Keep long sidebar usage reports recognizably sheet-sized while allowing the content to scroll. Short menu sheets retain dynamic sizing.
2026-10-01 04:55:17 +01:00
Mohamed Boudra 695ecb2812 fix(muse): make package installs and project rules work by default (#5775)
* fix(muse): make launch defaults usable and configure each agent

Disable sandboxing and trust workspaces by default so package installs and project rules work. Replace the host settings UI with validated per-agent providerOptions, preserving posture across session resume.

* docs(muse): describe shared provider options configuration
2026-09-30 20:09:10 +01:00
Mohamed Boudraandmillerben95 8332f4d1ff feat(muse): add Muse Code as a built-in provider plugin (#5719)
* test(muse): record real host protocol fixtures

* feat(muse): add built-in MSP provider

Co-authored-by: millerben95 <millerben95@users.noreply.github.com>

* fix(muse): expose provider settings in the app

The daemon registered host settings without a client contribution, leaving sandbox, network, and workspace trust unreachable. Share the definition with a settings screen and expose it through Command Center even when installed plugins are disabled.

* docs(muse): publish setup and provider support

* test: correct runtime fixtures and restart readiness

* fix(muse): hide reminder housekeeping from the timeline

* fix(muse): use the official Muse Code mark

Source: https://muse.ai/images/favicon/app-squiggle.svg

Preserve the official silhouette path verbatim. Remove its gradient definition and use currentColor, scaling the original 100-unit coordinates into the 24-unit viewBox. Black silhouette renders match pixel-for-pixel.

* fix(muse): adapt to current plugin APIs

Use the screen API introduced with sidebar plugin items. Usage reports now come from source account discovery; remove the retired agent usage-reference request and headline marker, and exercise discovery through the existing provider fixtures.

---------

Co-authored-by: millerben95 <millerben95@users.noreply.github.com>
2026-09-30 14:56:41 +01:00