Skip to content

fix(web): keep hidden models hidden after a harness switch - #24564

Merged
KiroCrewCaptain merged 1 commit into
mainfrom
crew/model-picker-hidden-ids
Oct 28, 2026
Merged

KiroCrewCaptain merged 1 commit into
mainfrom
crew/model-picker-hidden-ids

Conversation

@buluoray

@buluoray buluoray commented Oct 26, 2026 •

Copy link
Copy Markdown
Contributor

Problem / Motivation

Goal: a model hidden in Selectable Models stays hidden in the chat picker after switching Agent Harness.

Models a user hid in Settings > Chat > Model > Selectable Models came back in the chat model picker after switching Agent Harness, and the Selectable Models checkboxes no longer matched what they saved. The picker and the Settings list now compare hidden ids by model, not by exact string, so the same model stays hidden whichever harness names it.

Hidden id saved Picker row (other harness) Before After
claude-opus-4.7 (Kiro CLI) claude-opus-4-7[1m] (Claude Code) shown hidden
claude-opus-4.7 global.anthropic.claude-opus-4-7[1m] shown hidden
claude-opus-4.7 opus-4.7-1m (registry key) shown hidden
claude-opus-4.8 (1M) claude-opus-4-8 (200K, a different model) shown shown
opus (Claude Code bare alias) claude-opus-4.8 shown shown

Why it matters

A user who curated the picker loses that choice on every harness switch, and the Settings checkboxes stop telling the truth about what is saved.

Not a goal

  • Rewriting or migrating the stored dashboard.model_picker_hidden_models list.
  • Any backend change to how harnesses name models.

What changed (motivation → approach → change)

Root cause. dashboard.model_picker_hidden_models is one list of model ids for every harness, but each harness spells the same model its own way:

  • The Kiro backend lists kiro-cli ids, e.g. claude-opus-4.7 (src/kiro_crew/dashboard/handlers/agents.py, api_models).
  • The Claude Code backend lists registry keys or its own wire ids, e.g. claude-opus-4-7[1m] (src/kiro_crew/dashboard/handlers/agents.py, _cc_models).

The frontend compared those ids as raw strings: the chat picker in website/src/hooks/useInteractiveModels.ts (new Set(hiddenModels) then hidden.has(model.name)), and the Settings checkboxes in website/src/pages/settings/useChatPanelModels.ts. So after a switch no saved id matched, every hidden model reappeared, and the checkboxes showed the new harness's ids as selected.

  • website/src/hooks/useInteractiveModels.ts: new hiddenModelMatcher and hiddenIdsFor. They compare ids by normalizeModelKey (website/src/lib/model.ts), the canonical fold the dashboard already uses for "same model?" (picker, slot display, subagent downgrade flag). filterInteractiveModels uses it. A bare family alias (letters only: opus, sonnet, fable) is the exception and matches only itself, because it names whatever model the backend currently serves under it while the registry folds it to one fixed row.
  • website/src/pages/settings/useChatPanelModels.ts: the checkboxes use the same matcher. Showing a model removes every stored spelling of it, so an id saved under another harness cannot keep it hidden. Hidden ids for models not advertised now are still kept.

No backend change. The stored list is not rewritten on read.

Backwards compatibility

Compatible: the config key, its stored format and the API are unchanged. Ids the registry does not list (GPT, DeepSeek, operator-typed ids) fold losslessly (trim, lowercase, . to -), as before in normalizeModelKey. One behavior change: two ids that fold to the same canonical key now hide each other, which is the point of the fix; distinct registry models (200K vs 1M) stay distinct. Bare family aliases (opus, sonnet, fable) keep exact matching (case-insensitive): the registry maps opus to Opus 4.8 (1M), but Claude Code's opus is its newest Opus (#22985 labels that row "Opus 5.5"), so folding would hide the wrong model.

Removes nothing: the picker shows the same models it did for an unswitched harness; only cross-harness spellings of a hidden model are now hidden too.

Tests

  • website/src/test/useInteractiveModels.test.ts: 3 new tests. A model hidden under the kiro id stays hidden under three Claude Code spellings; claude-opus-4-8 (200K) and claude-opus-4.8 (1M) stay distinct; hiding opus does not hide Opus 4.8 or 5.5 rows, and hiding Opus 4.8 does not hide opus.
  • website/src/test/ChatPanel.modelVisibility.test.tsx: 1 new test. A model hidden under the other harness's spelling shows unchecked in Settings, and checking it removes both spellings from the saved list.
  • Before the fix: 2 of the 3 new tests failed (the "distinct models" one passes both ways by design). After: all pass. The bare-alias test fails without the bare-alias rule and passes with it.
  • The model-picker test files (28 files): all pass. tsc -p tsconfig.app.json: clean; eslint on the changed files: clean.

Proof: not provable -- frontend test (prove.py runs Python only)

Manual verification

Unit tests only (vitest + jsdom). Not checked in a real dashboard with a real harness switch, and not on the reporter's setup.

Screenshots / video

None yet: the change was checked with unit tests only, not in a running dashboard. The visible effect is which rows the chat model picker lists after a harness switch (table above).

Not captured on the author's host: an isolated pod cannot start there (the systemd user bus is not reachable from the agent sandbox), and a loopback dashboard needs a minted token and a browser approval that an unattended run cannot give. A maintainer with a desktop build can check it by hiding a Claude model on Kiro CLI, switching to Claude Code, and opening the picker.

Related Issues

No issue found. Searched open and closed issues and PRs for "hidden models reappear", "model_picker_hidden_models", "Selectable Models harness", "hidden model picker switch". Related, not the same: #17963 (MERGED, accepts qualified model ids in the visibility list), #22985 (OPEN, touches 3 of the same 4 files; a three-way merge with this branch had no conflict, and either can land first).

Source report: ethbradw, 2026-10-26.

Sibling sweep

Swept: git grep -nE "hiddenModels|model_picker_hidden" -- website/src
Sibling: website/src/components/useChatPanePickers.ts -- not affected -- it fetches the hidden list and passes it to useInteractiveModels, which does the matching.

Pattern harvest

Rule candidate: review-prompt
Pattern: comparing model ids from different harnesses as raw strings instead of through normalizeModelKey.

Checklist

  • At most two commits (one is the norm), with a Conventional Commits title
  • Existing tests pass and new tests added for new functionality
  • Self-review completed; code follows project style guidelines
  • Documentation updated (if applicable)
  • No secrets, credentials, or internal references in the diff

@buluoray
buluoray requested review from a team and KiroCrewCaptain as code owners October 26, 2026 20:41
@buluoray
buluoray requested a review from dodgeblaster October 26, 2026 20:41
@github-actions github-actions Bot added the readiness: checking Automated validation is still running label Oct 26, 2026
@github-actions

github-actions Bot commented Oct 26, 2026 •

Copy link
Copy Markdown
Contributor

UX Review (Opus 5.5) — ✅ PASS

UX-level review of d5f393a71a670c0272dfb0b0582f2a365b61e148 — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

UX-Verdict: PASS

Nothing blocks: the diff adds no control, string or layout. It only changes which existing picker rows and Selectable Models checkboxes count as hidden after a harness switch, so the missing blind read leaves no control unexamined.

Notes

  • The blind read did not run because the PR attaches no image. The fix is unit-tested only, so a before/after still of the chat model picker after switching from Kiro CLI to Claude Code would be the first proof that the effect shows up in a real dashboard.

[UX-REVIEWED] d5f393a

Description read

[DESCRIPTION-READ] 18f555591b87e394b8399b9bf059f8561ff8b33b5078aef6cd548c13e2ea3722

That sha256 names everything this verdict read: the PR description, plus the 3 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with .github/scripts/pr-description-capture.sh over the same inputs; if it differs, the description or that evidence changed after this verdict was formed and any finding drawn from either is unproven.

@github-actions

github-actions Bot commented Oct 26, 2026 •

Copy link
Copy Markdown
Contributor

Design Review (Opus 5.5) — ✅ PASS

Design-level review of d5f393a71a670c0272dfb0b0582f2a365b61e148 — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

Design-Verdict: PASS

Nothing blocks. Hidden ids are now matched through normalizeModelKey, the canonical fold the dashboard already uses, and bare aliases are kept exact. Ids don't change, so a revert undoes this cleanly.

Notes

  • The description says "3 new tests" and "1 new test", but the diff adds 4 and 2: the active-pin-respelled case and the select-all case are not mentioned. Update the Tests section so it matches the diff.
  • Only a filtering rule changes here and no new surface renders, so the missing-screenshot rule doesn't apply. The manual harness-switch check the description proposes is still the only end-to-end proof.

[DESIGN-REVIEWED] d5f393a

Description read

[DESCRIPTION-READ] 82efbbdf8db8e6e8e2e6d2793540c5d1157482f03e26d04dfa90293371b925a1

That sha256 names everything this verdict read: the PR description, plus the 1 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with .github/scripts/pr-description-capture.sh over the same inputs; if it differs, the description or that evidence changed after this verdict was formed and any finding drawn from either is unproven.

@github-actions

github-actions Bot commented Oct 26, 2026 •

Copy link
Copy Markdown
Contributor

First Principles Review (Opus 5.5) — ✅ PASS

Premise-level review of d5f393a71a670c0272dfb0b0582f2a365b61e148 — why this exists and whether the shipped surface is the smallest honest version. Updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

First-Principles-Verdict: PASS

The fix is at mechanism level: it reuses the existing normalizeModelKey fold. No raw-string comparison of the hidden list is left among the 16 files that reference it.

What this change ships

Inventory (8 items) — 5 justified

Intent: a model hidden in Selectable Models stays hidden in the chat picker whichever harness names it. FIX.

  1. The chat picker hides a hidden model under any harness's spelling of its id — justified
  2. In Settings, a model hidden under another harness's spelling now shows unchecked — justified
  3. Checking a model removes every stored spelling of it — justified
  4. "Select all" also removes the other harnesses' stored spellings — undeclared, rides along (needed for item 3's guarantee)
  5. "Deselect all" keeps every stored spelling in the saved list — undeclared, rides along
  6. A bare family alias (opus, sonnet) matches only itself — justified
  7. An active session pin stays visible when the harness spells it differently — undeclared, rides along (keeps the pin visible now that hiding matches by key)
  8. Hidden ids for models not advertised now are still detected and kept, matched by model — justified

[FIRST-PRINCIPLES-REVIEWED] d5f393a

Description read

[DESCRIPTION-READ] 8c29be4ff682597f454cc88c3a690ec28038528d7378b8f69512c3303b8f578f

That sha256 names the PR description this verdict read. Recompute it with .github/scripts/pr-description-capture.sh; if it differs, the description changed after this verdict was formed and any finding drawn from the description is unproven.

@github-actions

github-actions Bot commented Oct 26, 2026 •

Copy link
Copy Markdown
Contributor

Opus 5.5 Review — ✅ no blocking findings

Reviewed d5f393a71a670c0272dfb0b0582f2a365b61e148 — this comment is updated in place on each push.

Review details

No findings.

[OPUS-REVIEWED] d5f393a

Verdict parsed from the review's SHA-scoped output markers for commit d5f393a71a670c0272dfb0b0582f2a365b61e148.

False positive or not applicable? A repository writer can comment:
/ai-review override fable d5f393a71a670c0272dfb0b0582f2a365b61e148 false-positive: <reason, 20 to 500 characters>

@github-actions

github-actions Bot commented Oct 26, 2026 •

Copy link
Copy Markdown
Contributor

GPT 6.1 Review — ✅ no blocking findings

GPT 6.1 completed its review of d5f393a71a670c0272dfb0b0582f2a365b61e148 and found no blocking issues.

This comment is updated in place on each push.

Review details

No findings.
[GPT-REVIEWED] d5f393a

False positive or not applicable? A repository writer can comment:
/ai-review override gpt d5f393a71a670c0272dfb0b0582f2a365b61e148 false-positive: <reason, 20 to 500 characters>

@github-actions github-actions Bot added readiness: action required A blocking check or review needs attention and removed readiness: checking Automated validation is still running labels Oct 26, 2026
@buluoray
buluoray force-pushed the crew/model-picker-hidden-ids branch from 1ef3e5b to 810fcf4 Compare October 26, 2026 21:04
@buluoray

Copy link
Copy Markdown
Contributor Author

self-added: yes
mechanism: Select all / Deselect all deltas for cross-harness hidden spellings

  • span=ba5c17b4d86d Select all leaves cross-harness hidden IDs persisted

fixed in 810fcf4: selectAllModels sends every stored spelling of an advertised model in remove, as toggleVisibleModel does; deselectAllModels keeps the full stored list in its optimistic next, matching the server's append-only add.
New test select all removes a model hidden under another harness spelling fails on the previous head and passes now.

@buluoray

Copy link
Copy Markdown
Contributor Author

self-added: yes
mechanism: Select all / Deselect all deltas for cross-harness hidden spellings

  • span=4731c7e4cd66 selectAllModels removes only advertised ids

fixed in 810fcf4: remove is the union of the advertised ids and hiddenModels.filter(isAdvertisedModel); deselectAllModels builds next from the full current hiddenModels plus the added ids.
New test select all removes a model hidden under another harness spelling fails on the previous head and passes now.

@buluoray
buluoray force-pushed the crew/model-picker-hidden-ids branch from 810fcf4 to a28860a Compare October 26, 2026 21:34
@github-actions github-actions Bot added readiness: checking Automated validation is still running and removed readiness: action required A blocking check or review needs attention labels Oct 26, 2026
@buluoray

Copy link
Copy Markdown
Contributor Author

self-added: yes
mechanism: the active-model keep set in filterInteractiveModels compared raw ids while hidden matching used canonical keys

  • span=c261403ee26a Preserve active models across harness spellings

fixed in a28860a: filterInteractiveModels now keeps a row when its id equals an active pin or its hiddenModelKey equals an active pin's key, so a hidden pin respelled by the other harness (claude-opus-4.8 vs claude-opus-4-8[1m]) keeps its row.
New test keeps a hidden active model visible when the harness spells it differently fails on 810fcf4 and passes now; the 6 picker test files (92 tests), tsc -p tsconfig.app.json and eslint are green.

@github-actions github-actions Bot added readiness: action required A blocking check or review needs attention and removed readiness: checking Automated validation is still running labels Oct 26, 2026
…its id

dashboard.model_picker_hidden_models is one list of model ids, but each harness
names the same model its own way: the kiro backend lists kiro-cli ids
(claude-opus-4.7), the Claude Code backend lists registry keys or its own
wire ids (claude-opus-4-7[1m]). The picker and Settings > Selectable Models
compared raw strings, so after a harness switch every hidden model came back
and the checkboxes no longer matched what the user saved.

Compare hidden ids by normalizeModelKey, the canonical fold the dashboard
already uses for "same model?". Showing a model in Settings removes every
stored spelling of it. Distinct registry models (claude-opus-4-8 200K vs
claude-opus-4.8 1M) stay distinct.
@buluoray
buluoray force-pushed the crew/model-picker-hidden-ids branch from a28860a to d5f393a Compare October 27, 2026 22:14
@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: passed Eligible automated validation passed for the current revision and removed readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running labels Oct 27, 2026
@KiroCrewCaptain
KiroCrewCaptain enabled auto-merge (squash) October 28, 2026 00:05

@KiroCrewCaptain KiroCrewCaptain left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Tier 1 auto-approve: fix (4 files). Criteria: no conflict, no requested changes, security path denylist clean, design-doc gate clean, security checklist all-NO, no reviewer-lane findings, readiness signal fresh for this head SHA, declared category matches the diff. Category: the hidden-model list is matched by canonical model key, so a model hidden under one harness stays hidden after a harness switch; covered by new tests.

@KiroCrewCaptain
KiroCrewCaptain merged commit 501f455 into main Oct 28, 2026
109 of 110 checks passed
@KiroCrewCaptain
KiroCrewCaptain deleted the crew/model-picker-hidden-ids branch October 28, 2026 00:06
@github-actions github-actions Bot removed the readiness: passed Eligible automated validation passed for the current revision label Oct 28, 2026
bolichen97 added a commit that referenced this pull request Oct 29, 2026
Cherry-picks these main commits onto release/0.11.0, in merge order:

- fix(dashboard): tell the crewmate about unresolved fields, draw no stale band (#24069)
  (cherry picked from commit f9d8e19)
- fix(mcp-gateway): never pool a credential vendor (#24672)
  (cherry picked from commit 2464225)
- fix(acp): keep MCP tools direct on KAS when the spec grants no tool_search (#24297)
  (cherry picked from commit ddc1fea)
- fix(crew): move a working icon avatar as one still image so it cannot tear (#24678)
  (cherry picked from commit e062153)
- fix(crew-mode): keep a crewmate's own chat on the conductor (#24894)
  (cherry picked from commit 7582829)
- fix(memory): name a damaged memory.db instead of restore advice (#24567)
  (cherry picked from commit efe7624)
- fix(env): search Builder Toolbox's Windows bin dir for agent binaries (#24563)
  (cherry picked from commit 686d39a)
- fix(discord): re-admit bot-opened threads after a gateway restart (#21776)
  (cherry picked from commit 072a37b)
- fix(web): keep a hidden model hidden after a harness switch respells its id (#24564)
  (cherry picked from commit 501f455)
- fix(gateway): say the gateway is running once boot finishes (#25029)
  (cherry picked from commit 02627ec)
- fix(mcp): say why an mcp.json settings file could not be read (#22169) (#24572)
  (cherry picked from commit ed75593)
bolichen97 added a commit that referenced this pull request Oct 29, 2026
Cherry-picks these main commits onto release/0.11.0, in merge order:

- fix(dashboard): tell the crewmate about unresolved fields, draw no stale band (#24069)
  (cherry picked from commit f9d8e19)
- fix(mcp-gateway): never pool a credential vendor (#24672)
  (cherry picked from commit 2464225)
- fix(acp): keep MCP tools direct on KAS when the spec grants no tool_search (#24297)
  (cherry picked from commit ddc1fea)
- fix(crew): move a working icon avatar as one still image so it cannot tear (#24678)
  (cherry picked from commit e062153)
- fix(crew-mode): keep a crewmate's own chat on the conductor (#24894)
  (cherry picked from commit 7582829)
- fix(memory): name a damaged memory.db instead of restore advice (#24567)
  (cherry picked from commit efe7624)
- fix(env): search Builder Toolbox's Windows bin dir for agent binaries (#24563)
  (cherry picked from commit 686d39a)
- fix(discord): re-admit bot-opened threads after a gateway restart (#21776)
  (cherry picked from commit 072a37b)
- fix(web): keep a hidden model hidden after a harness switch respells its id (#24564)
  (cherry picked from commit 501f455)
- fix(gateway): say the gateway is running once boot finishes (#25029)
  (cherry picked from commit 02627ec)
- fix(mcp): say why an mcp.json settings file could not be read (#22169) (#24572)
  (cherry picked from commit ed75593)
iamwhatever pushed a commit that referenced this pull request Oct 29, 2026
Cherry-picks these main commits onto release/0.11.0, in merge order:

- fix(dashboard): tell the crewmate about unresolved fields, draw no stale band (#24069)
  (cherry picked from commit f9d8e19)
- fix(mcp-gateway): never pool a credential vendor (#24672)
  (cherry picked from commit 2464225)
- fix(acp): keep MCP tools direct on KAS when the spec grants no tool_search (#24297)
  (cherry picked from commit ddc1fea)
- fix(crew): move a working icon avatar as one still image so it cannot tear (#24678)
  (cherry picked from commit e062153)
- fix(crew-mode): keep a crewmate's own chat on the conductor (#24894)
  (cherry picked from commit 7582829)
- fix(memory): name a damaged memory.db instead of restore advice (#24567)
  (cherry picked from commit efe7624)
- fix(env): search Builder Toolbox's Windows bin dir for agent binaries (#24563)
  (cherry picked from commit 686d39a)
- fix(discord): re-admit bot-opened threads after a gateway restart (#21776)
  (cherry picked from commit 072a37b)
- fix(web): keep a hidden model hidden after a harness switch respells its id (#24564)
  (cherry picked from commit 501f455)
- fix(gateway): say the gateway is running once boot finishes (#25029)
  (cherry picked from commit 02627ec)
- fix(mcp): say why an mcp.json settings file could not be read (#22169) (#24572)
  (cherry picked from commit ed75593)
bolichen97 added a commit that referenced this pull request Oct 29, 2026
Parts 2 and 3 of the 0.11.0 list, combined after part 3 (#25666) was merged into part 2's branch.

Cherry-picks these main commits onto release/0.11.0, in merge order:

- fix(desktop): say why a pop-out cannot open instead of blaming the browser (#23922)
  (cherry picked from commit 99c2474)
- fix(crew): list a crewmate's own script and command schedules (#24023)
  (cherry picked from commit d2d8d14)
- fix(mcp-gateway): show a refused share outside Developer Mode (#23917)
  (cherry picked from commit c612613)
- fix(apps): negative-cache failed app.json fetches in the store listing (#23925)
  (cherry picked from commit 455e8ec)
- fix(power): say that Windows Modern Standby can still sleep on lock (#18259)
  (cherry picked from commit 0480058)
- fix(chat): show nothing in the transcript when the context cap trims (#24492)
  (cherry picked from commit 1411fcc)
- fix(frontend): seed settings rows from the served config during render (#24625)
  (cherry picked from commit e198b63)
- fix(schedule): keep the empty-folder chip rename open and return focus after it (#23904) (#24632)
  (cherry picked from commit 9f51b91)
- feat(website): swipe to dismiss the notification banner on mobile (#24674)
  (cherry picked from commit e4fad39)
- fix(agent): drop a gateway-inherited scratch temp before an agent spawn (#24645)
  (cherry picked from commit 1f28708)

- fix(dashboard): tell the crewmate about unresolved fields, draw no stale band (#24069)
  (cherry picked from commit f9d8e19)
- fix(mcp-gateway): never pool a credential vendor (#24672)
  (cherry picked from commit 2464225)
- fix(acp): keep MCP tools direct on KAS when the spec grants no tool_search (#24297)
  (cherry picked from commit ddc1fea)
- fix(crew): move a working icon avatar as one still image so it cannot tear (#24678)
  (cherry picked from commit e062153)
- fix(crew-mode): keep a crewmate's own chat on the conductor (#24894)
  (cherry picked from commit 7582829)
- fix(memory): name a damaged memory.db instead of restore advice (#24567)
  (cherry picked from commit efe7624)
- fix(env): search Builder Toolbox's Windows bin dir for agent binaries (#24563)
  (cherry picked from commit 686d39a)
- fix(discord): re-admit bot-opened threads after a gateway restart (#21776)
  (cherry picked from commit 072a37b)
- fix(web): keep a hidden model hidden after a harness switch respells its id (#24564)
  (cherry picked from commit 501f455)
- fix(gateway): say the gateway is running once boot finishes (#25029)
  (cherry picked from commit 02627ec)
- fix(mcp): say why an mcp.json settings file could not be read (#22169) (#24572)
  (cherry picked from commit ed75593)
bolichen97 added a commit that referenced this pull request Oct 29, 2026
Backports parts 2-4 of the 0.11.0 insider list (#25665, #25666, #25667) onto
release/0.11.0. Parts 0-1 landed in e379f74 (#25663, #25664).

- fix(desktop): say why a pop-out cannot open instead of blaming the browser (#23922)
  (cherry picked from commit 99c2474)
- fix(crew): list a crewmate's own script and command schedules (#24023)
  (cherry picked from commit d2d8d14)
- fix(mcp-gateway): show a refused share outside Developer Mode (#23917)
  (cherry picked from commit c612613)
- fix(apps): negative-cache failed app.json fetches in the store listing (#23925)
  (cherry picked from commit 455e8ec)
- fix(power): say that Windows Modern Standby can still sleep on lock (#18259)
  (cherry picked from commit 0480058)
- fix(chat): show nothing in the transcript when the context cap trims (#24492)
  (cherry picked from commit 1411fcc)
- fix(frontend): seed settings rows from the served config during render (#24625)
  (cherry picked from commit e198b63)
- fix(schedule): keep the empty-folder chip rename open and return focus after it (#23904) (#24632)
  (cherry picked from commit 9f51b91)
- feat(website): swipe to dismiss the notification banner on mobile (#24674)
  (cherry picked from commit e4fad39)
- fix(agent): drop a gateway-inherited scratch temp before an agent spawn (#24645)
  (cherry picked from commit 1f28708)

- fix(dashboard): tell the crewmate about unresolved fields, draw no stale band (#24069)
  (cherry picked from commit f9d8e19)
- fix(mcp-gateway): never pool a credential vendor (#24672)
  (cherry picked from commit 2464225)
- fix(acp): keep MCP tools direct on KAS when the spec grants no tool_search (#24297)
  (cherry picked from commit ddc1fea)
- fix(crew): move a working icon avatar as one still image so it cannot tear (#24678)
  (cherry picked from commit e062153)
- fix(crew-mode): keep a crewmate's own chat on the conductor (#24894)
  (cherry picked from commit 7582829)
- fix(memory): name a damaged memory.db instead of restore advice (#24567)
  (cherry picked from commit efe7624)
- fix(env): search Builder Toolbox's Windows bin dir for agent binaries (#24563)
  (cherry picked from commit 686d39a)
- fix(discord): re-admit bot-opened threads after a gateway restart (#21776)
  (cherry picked from commit 072a37b)
- fix(web): keep a hidden model hidden after a harness switch respells its id (#24564)
  (cherry picked from commit 501f455)
- fix(gateway): say the gateway is running once boot finishes (#25029)
  (cherry picked from commit 02627ec)
- fix(mcp): say why an mcp.json settings file could not be read (#22169) (#24572)
  (cherry picked from commit ed75593)

- fix(mcp-gateway): render a refused share through ErrorNotice (#25660)
  (cherry picked from commit d895a90)
- fix(crew): write a crewmate's hidden welcome and check-ins in the user's language (#24966)
  (cherry picked from commit d5a76e6)
- feat(crew-mode): move every used crewmate to its conductor on upgrade (#24954)
  (cherry picked from commit 76e89ae)
- fix(hooks): Reads mode auto-approves Kiro Crew's own read-only MCP tools (#24910)
  (cherry picked from commit d8f2038)
- fix(members): list crewmates the person created, even with an empty chat (#24987)
  (cherry picked from commit 163652b)
- feat(cron): let a script cron feed its crewmate's Dashboard page (#24692)
  (cherry picked from commit 959b35a)
- fix(kas): honour a server's autoApprove through the derived permissions (#25106)
  (cherry picked from commit 073c83a)
- fix(chat): keep a scrolled-up reader still when a reprice clamps them (#24591)
  (cherry picked from commit 9a73292)
- fix(security): restore every allowed link on a host, not only identical ones (#20487)
  (cherry picked from commit b161144)
- fix(website): swipe the mobile notification banner up to dismiss it (#25151)
  (cherry picked from commit e68a716)
- test: carry the _card_outcomes helper #24910's tests read
  (cherry picked from commit 9df4d01)

The #24987 decision entry is new on this branch, so its Evidence lines
now link the owner restatement and the maintainer approval on #24987
instead of describing them. Main keeps its shipped (immutable) copy.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants