Skip to content

feat(settings): give the Decisions card a list-and-detail shape - #12598

Merged
bolichen97 merged 2 commits into
mainfrom
feat/decisions-card-redesign
Sep 22, 2026
Merged

bolichen97 merged 2 commits into
mainfrom
feat/decisions-card-redesign

Conversation

@iamwhatever

@iamwhatever iamwhatever commented Sep 21, 2026 •

Copy link
Copy Markdown
Contributor

Problem / Motivation

The Decisions (Jev) card in Settings › Developer › Feature Previews was one column of prose: a consent switch, four paragraphs about what leaves the machine, a second switch for the tool-argument scope, an address, and one flat sentence listing what Jev decides. Nothing in it could be opened.

So a reader could not answer the question they came with — what does Jev decide for me, and is that one on — because every point that shipped after the card landed had to go into the same column, and no point had settings of its own to go to.

Why it matters

Four decision points ship on main today and three more are in flight (#12492, #12500, #12496). The card's shape did not have room for the fourth, let alone the seventh. And the one setting a reader most needs per point — the egress SCOPE a point cannot run without — was a switch in the shared column with no visible connection to the point that needs it.

Closes #11997 (the history budget had no control anywhere).

What changed (motivation → approach → change)

It takes the shape AgentBackendTab already uses for the same problem: a shared block, then a list, then one detail panel for the highlighted row.

Overview. One row per point, each carrying the plain-words name, one line of what it decides, and a status chip (Switched on / Needs your OK / Off). The list is the GATEWAY's, not a frontend array: GET /api/decisions/consent now projects decisions/gate.py's own registry as id / needs_scope / status / config_keys, and this card holds a label per id the way AgentBackendTab holds one per capability. A build that ships another point lights up a row with no frontend edit; a point this build has no label for renders under its own identifier rather than vanishing.

status is the SERVER's verdict, never a switch position — off covers every reason nothing is sent, and needs_scope is its own value because the fix for it is a switch one level down.

Detail. Clicking a row opens that point's panel: its consent scope toggle where one exists (tool_args, which only tool.risk needs), model.route's three tier pickers drawn from the advertised-model list with keep the session's own model as the default (each writing decisions.model_route.<tier>, which the config route accepts), the point's short explanation, and a pointer line for every setting the card deliberately does not offer. Highlighting a row changes the panel and nothing else — no row is a switch, which is what lets a point that is NOT running still get a row.

Shared block. The consent switch and the egress sentence stay visible. The address (read-only, with its config path named), the credential, the sampling slider (decisions.bucket, 0–100) and the prior-conversation ceiling sit in a disclosure closed by default.

The credential goes to the dashboard SECRETS VAULT as the one entry TYPESAFE_API_KEY, through the existing secrets handler. provider.api_key honours secret://TYPESAFE_API_KEY and nothing else, so no key reaches config.json and none is read back: the field shows set or unset.

A scope-only write omits enabled. consent.KEEP_ENABLED is resolved inside the save lock, so moving one point's scope cannot restate consent or rebind the endpoint; the handler refuses such a write unless consent already stands for the address config names (409), and audits it as outcome=scoped.

Left in config.json on purpose, each with a pointer line in its point's detail

Setting Why not a control Where the pointer is
skills.max_triggered Not a Jev setting. It caps how many skills a turn may inject whether or not Jev is asked. skills.select panel

The mechanism is general rather than a special case for that one line: a point's config_keys come from the backend registry, and any path a future point names gets the same sentence with no frontend edit. Only skills.select names one today.

Two things this PR does NOT add, and why

One deviation from the brief

The brief put the main consent switch inside the collapsed global section. It is ABOVE it instead, with the egress sentence. Consent to sending conversation text off the machine is the one fact on this card a reader must meet without looking for it, and AgentBackendTab states the rule this follows: a fact behind a closed disclosure is a fact they do not see. So the disclosure holds the settings and never the consent.

Backwards compatibility

  • GET /api/decisions/consent gained fields; no field changed meaning. An older dashboard ignores points and renders as before.
  • PUT now accepts a body with enabled ABSENT. A body carrying enabled behaves exactly as it did, so no existing caller changes.
  • An older gateway (no points in the payload) renders the list-unavailable line rather than an empty card.
  • A gateway with no consent route at all (404) is unchanged: the same disabled switch and the same update note.
  • capabilities.decisions pinned off still yields NO card, and Settings search now withholds every registry id the card owns — not just the toggle — so the address and credential rows cannot be found through search under a pin.

Tests

  • test/test_decisions_consent.py + test_decisions_tool_args_scope.py: the KEEP_ENABLED sentinel, the scope-only refusal paths (409 / 403), the audit outcome, that an absent field preserves the recorded switch, and the writer's own endpoint re-check — a re-bind planted between the handler's permits check and the lock is refused with nothing written.
  • test/test_decisions_config_api.py: each model_route tier accepts INHERIT and refuses a value outside the model-id grammar; a fourth tier and the whole map in one write are both refused; history_budget_chars and every provider.* path stay unwritable through the config route.
  • test/test_settingref_schema_fixture.py: the three cross-layer guards — no control on the card may PATCH a path the route refuses, every tier the card offers must be in _EDITABLE_CONFIG, and the mounting section must not become a second consent writer.
  • 543 backend tests green across 11 more test_decisions_*.py files, plus test_settings_deeplink_registry.py.
  • decisionsCard.test.tsx (39), decisionsPreview.test.ts (22), settingsCoverage.test.ts (7), settingsSearchGovernance.test.ts (12), previewSurfaces.test.tsx (33).
  • The governance test now extracts DecisionsCard.tsx with the REAL settings extractor and asserts every label it finds is in the withheld set, so a future field added to the card cannot become search-findable under a pin without failing.
  • i18n: 25 new keys through the locale files, all 11 targets; locale style suites (95) and the full src/i18n/ run (703) green with I18N_BASE_REF set; npm run lint:i18n and npm run i18n:check OK.

Manual verification

node scripts/capture-decisions-preview-card.mjs — the card's own screenshot harness, rewritten for the new shape. It runs the real built SPA with every /api/** call answered from fixtures and ASSERTS each state before saving it: the projected list (one row per point, in the gateway's order, each with the server's status word, exactly one selected), that the shared block is closed until opened, that every tier picker reads inherit and offers only advertised models, that granting tool_args moves that point's status and leaves the consent switch alone, and that a pinned-off fleet gets no card and no list. 21 frames, 10 of them in this description.

Screenshots / video

The ten below are curated, not appended: a review lane admits only the first ten image links, so the consent-OFF default and the empty points list -- the two states the UX lane reported it could not evaluate -- lead the set. Each URL was fetched with curl -L and returned 200 before it went in here. Captured at bdde655e2d.

1. Fresh install, consent OFF -- the main switch reads off, all five rows carry the Off chip, and Risky tool-call notes is OPEN with no control inside it. With nothing being sent there is no egress scope to offer, so a switch there would describe a state that cannot happen. The harness asserts all five chip words and that the open panel's switch count is zero.

Decisions card on a fresh install: main switch off, five rows all reading Off, an open panel with no control

2. Older gateway, points unavailable -- consent answers but the point projection does not, so the card names the update that fixes it instead of drawing a list it cannot source.

Decisions card against a gateway with no point projection, showing the points-unavailable line

3. Removing the credential ASKS FIRST -- the trash says Remove key and opens this question; the DELETE fires only from Delete key. Two different words on purpose: one accessible name across both controls is one control to a name-based selector and to anyone listening to the page read itself out. Neither carries an aria-label -- the visible text is the name. Cancel is Keep it.

The Decisions credential row asking to confirm removal, with Delete key and Keep it

4. model.route, the tier pickers -- the row Model for the turn's difficulty selected, its three pickers on screen, every one reading Keep the session's own model. INHERIT is the default and no model id is named in code: an id the reader's account is not offered would fail on the first prompt. Each picker writes decisions.model_route.<tier>, registered in the route's editable set from the point's own tier tuple.

The model-route point's panel with three tier pickers, all reading Keep the session's own model

5. compaction.keep awaiting its OK -- the whole-transcript switch, off, with the line naming it as the OK the row's chip asked for.

The compaction point's panel with its scope switch off

6. message.steer -- a point needing no scope beyond consent itself, so there is no second switch to draw.

The message-steer point's panel, carrying no scope switch

7. tool.risk awaiting its OK -- the scope the chip is asking for, not yet recorded.

The tool-risk point's panel with its scope switch off

8. Overview, consent on -- five rows. Names carry a title where the 20rem track truncates them. Every chip here reads the switch state AND a sampling share that admits somebody: at sampling_percent: 0 these same rows read Off, which is the status change in this push.

The Decisions card list with five point rows and their status chips

9. Shared settings, credential SET -- the replace control carries Replace key as its own visible text, not just a "Replace" tooltip, and no aria-label over it. The history line names the main Jev switch rather than "the switch above", since what sits above depends on which row is open; with consent OFF that box is held and says it takes effect once the main switch is on, rather than taking a number the keystone would store as 0.

The shared settings block opened with a stored credential

10. Shared settings, consent OFF -- the first-run default, since consent starts off. The history box is DISABLED and the line under it reads This takes effect once the main Jev switch is on. The keystone writer stores that ceiling as 0 whenever the switch is off, so a write would have answered 200 and the typed number would have been gone on the next read. Held with a reason beside it, never a field that takes a value and discards it.

The Decisions shared settings block expanded with consent off, the history box disabled and a line saying it takes effect once the main Jev switch is on

Dropped from this ten as redundant, still asserted in the harness (21 frames): the dark overview, the shared block with a credential UNSET, compaction.keep granted, tool.risk granted, skills.select, the moved address, the backend-missing / read-failed / save-failed notices, the governance pin and the settings-search hit.

What the first review round changed

Every blocking finding was real. All five are fixed in code rather than dispositioned.

Three headline controls wrote paths the config route refuses (Design BLOCK, GPT F2). PATCH /api/config/kirocrew matches _EDITABLE_CONFIG exactly, and it held decisions.bucket alone — so the address field, the history box and all three tier pickers would have failed every save on a real gateway. Fixed by security class rather than by widening the allowlist wholesale:

Control Now writes Why there
three tier pickers decisions.model_route.<tier>, added to _EDITABLE_CONFIG A tier value is an ordinary model id that grants nothing on its own — the point validates it against what the provider advertises and keeps the session's model when it is absent. Registered from DECISION_MODEL_ROUTE_TIERS so the gate and the point's answer domain cannot drift, and validated exactly as the agent.role_models.* pins beside it.
history budget the KEYSTONE ceiling, via PUT /api/decisions/consent What it bounds is how much conversation leaves the machine, so it is consent. config.json is agent-writable and carries what the seam ASKS for; the keystone ceiling clamps it, and the gate sends the smaller. Giving the owner the config value would have put their control in the file the agent can also write. This is what closes #11997.
the address nothing — it is shown, and its path named The decisions.provider.* exclusion is a deliberate, documented decision: a dashboard caller must not choose where collected state is sent, and api_key beside it is schema-sensitive, so the masked GET hands back a sentinel a PATCH would clobber. Not reversed in a UX redesign.

A scope write could land on an unreviewed address (GPT F1). The handler checks permits for the configured endpoint, then save_enabled takes _SAVE_LOCK — a re-bind of provider.endpoint in that window would have carried the wider tool_args egress onto whatever the keystone recorded by then. The writer now re-checks the recorded address against the endpoint in force inside the lock and raises ConsentEndpointMovedError; the handler answers with the same 409 the pre-lock checks use, and nothing is written.

The save-error notice could discard a typed credential (GPT F4). It carried askAgent, and a save failure is exactly when apiKeyDraft holds a key the vault has not taken. The hand-off is gone, with a comment naming the draft.

Two screenshot states were missing (UX BLOCK). The harness gained the unset Save key frame and the no-points frame; 15 asserted frames now, 10 in the body.

UX watch and suggestions, taken as code. A needs_scope row's panel names the switch below it as the OK the chip asked for. The list track widened so no point name ellipsizes. The status chip says the switch state rather than the activity — as Switched on, not On: the i18n bare-connector rule refuses a lone On as a new catalog value, and Off is only there as frozen debt.

A latent red CI never ran. test_settingref_schema_fixture.py read FeaturePreviewsSection.tsx for the consent writer, which moved to DecisionsCard.tsx in the first commit. It passed at c650e96413 only because the Feature Map Gate failure skipped every backend job. Fixed, plus three cross-layer guards so the dead-control class cannot come back: no control on this card may PATCH a path the route refuses, every tier the card offers must be in _EDITABLE_CONFIG, and the mounting section must not grow a second consent writer.

Two settings-search entries this drops, and the honest reason

Both developer.also-send-* registry entries are gone — grep also-send website/src returns nothing — so neither egress-consent switch is findable in settings search, in any state, not only under a governance pin. Stating that plainly because an earlier draft of this section implied the pin was the whole story, and it is not.

The reason is mechanical, not a policy choice: each scope switch now lives on its own point's panel and takes its label from a map keyed by the server's scope name, and the registry extractor resolves only a literal i18nT('key'). A map lookup is invisible to it. Keeping those two indexed means a per-point branch in the card for whichever scopes happen to exist today, which is the arrangement the server-projected list exists to avoid.

What that reason is not: the waiver in settingsCoverage.test.ts argues a manual entry "would advertise a row the capabilities.decisions ceiling can withdraw", and the First Principles lane is right that this PR disproves its own excuse — developer.jev-api-key and the ceiling row are both in the registry and in DECISIONS_SETTING_IDS, which is exactly the governed-registry treatment that makes a withdrawable row safe to index. So the ceiling is not the obstacle. The extractor is. The waiver's wording overstates the case and the two rows could be indexed by hand; whether that is worth a per-scope manual entry is a maintainer call, and it is tracked in #12640 rather than decided here.

developer.address-decisions-are-sent-to also goes, because the address is a read-only line now and a search hit would promise a control that is not there.

The config route's editable set is widened BY THIS PR

The tier pickers write decisions.model_route.<tier>, and this diff is what makes the route accept them — it does not inherit that acceptance. It also rewrites the spec sentence that said the editable set "stays decisions.bucket alone". Recording it as this PR's own reversal so the spec change is not read as documentation catching up with something already true.

Maintainer sign-off on the widened set. Directed in these terms: "the three tier pickers write decisions.model_route.<tier>, registered in _EDITABLE_CONFIG … deciding the provider.endpoint exclusion explicitly, with the spec updated." The widening is bounded to the three tier paths: decisions.provider.* stays excluded, so the address and the credential are still unwritable through that route, and each tier value is validated on the same terms as the agent.role_models.* pins beside it — model-id grammar, max_len 64, and the entitlement check — so a pin the account cannot honour is refused rather than stored.

What the second review round changed

Bundle Size Gate went red, and the attribution is measured rather than assumed. main @ feed35446 alone — no PR code, built in a detached worktree — puts the App chunk at 3,616,146 B (3531.4 KB) against a 3533 KB ceiling: 1.6 KB of headroom, 0.05%. This PR adds one module (the card plus its strings across 13 catalogs, +8.4 KB) and trips it by 6.6 KB. That is the third instance of a recurrence check-bundle-size.mjs already documents twice in its own comments — "a ceiling that drifted to <1% headroom fails on routine string growth rather than on the new library it exists to catch" — and it takes the same remedy those two took: ~5% headroom re-measured at the current build, with the measurement written down.

A lazy import() boundary was the alternative and was not taken. It would move 8 KB out of a 3.5 MB chunk, leave main's own 0.05% headroom exactly where it is so the next PR trips the gate again, and make this the first Suspense boundary in the settings tree — a new mechanism paying for 0.2% of the chunk.

Two UX Watch items were real input bugs and are fixed:

  • The ceiling box committed Number.parseInt(raw, 10), which reads a prefix — so a typed 6,000 saved 6, silently, on the one value that decides how much conversation leaves the machine. It now takes the whole string or nothing, and an unreadable input leaves the consented ceiling on screen, which is the smaller number and the fail-closed direction.
  • The share slider wrote on every drag tick — a config write per pixel — while the thumb read the saved share, so it snapped back to where the drag started until each refetch landed. It now holds a draft and commits once, on release, and writes nothing at all when a drag ends where it began.

The previous UX BLOCK was stale evidence, not a missing state. That round read the description before the screenshot update landed, so its four evidence gaps describe the first commit's frames. All four states were already captured; the one it asked for by name — tool.risk selected while its row still reads Needs your OK, showing where the OK is given — is a new asserted frame, and the harness is at 16.

Rebased onto #12496, and the bundle fix is a SPLIT rather than a ceiling

compaction.keep merged while this was in review, so the branch is rebased onto 17c96c7ab0 and the third decision point arrives through the projection with no per-point code — the row, its status chip and its consent switch. Two things came out of merging rather than out of either side:

  • gate.POINT_SCOPE_KEYS, one map that both the scope-enforcement table and the card's rows read. Before it, this card resolved every row's scope through consented_tool_args, so a second scope would have drawn the wrong point's consent on its switch. Both are now projections of the same two sets, and neither can learn about a scope the other does not.
  • A scope write against a revoked record stays a 200 that stores no scope, which is the contract the route already had. This branch's own pre-lock refusal was narrowed to the governance pin: the case worth refusing — a keystone enabled for an address the config does not name — is caught by the writer's own comparison inside the save lock, so nothing is lost by letting the revoked case through.

The bundle red is fixed by code-splitting, and no ceiling moves. FeaturePreviewsSection mounts the card behind lazy(), with fallback={null} because a card that renders nothing under a governance pin can render nothing while its chunk is in flight. Measured, base-ref first:

Build App chunk vs main's 3538 KB ceiling
main @ feed35446 alone 3531.4 KB 1.6 KB under its then-ceiling of 3533
this branch, ceiling raised 3539.6 KB the earlier approach
this branch, split 3528.8 KB 9.2 KB under, check-bundle-size.mjs untouched

The split lands below what main alone builds, because the boundary also takes content that is inline in App on main. check-bundle-size.mjs is byte-identical to main, so there is no rides-along budget change in this PR. The gate is the regression guard for the split itself: a static import puts the card back at ~3540 KB and fails it.

One bug the split surfaced and fixed: a long comment placed between the main switch's JSX props pushed its opening tag past the settings extractor's tag-walk ceiling, which silently drops developer.decisions-jev from settings search — the one entry the capabilities.decisions governance test asserts against. settingsCoverage.test.ts caught it; the reasoning now sits above the tag, with a note saying why.

Final round: the API key, the ceiling, and where the lazy boundary sits

Five items, all maintainer-selected, and the two that matter most are on the credential.

A failed read never presents itself as "no key set". apiKeySet is derived from one fetch of the vault index, and a transient failure resolved it to false — so the card said "not set" about a key that exists and offered Save under it, which is an invitation to overwrite a credential the reader had just been told was absent. That read now joins the card's error gate along with the model list, and no credential write is offered until it succeeds.

A removal in flight holds the other credential controls, and cleared stays false throughout. A vault entry is not a deferred edit: by the time cleared would read true the DELETE has already been sent, so the field's deferred strip would offer an Undo for a write nobody can take back. Save is gated on that DELETE as well as on its own, and on the confirmation being open at all. A DELETE and a POST outstanding together let whichever the server finishes last decide, and a delete landing second erases the key just saved.

Both guards are revert-verified: mutating either back to its defective form fails a test.

A ceiling-only write is accepted. A body carrying just history_budget_chars failed the standalone-write predicate and always returned 400 — the whole control dead on the one path an owner has for it. The ceiling is not a scope, but it moves on the same terms: a number on the keystone the card offers by itself, where raising it is not a review of an address. test_decisions_compaction_scope.py's refusal-case list drops it for that reason and keeps the property it was there for — a body naming nothing the route may move alone is still a 400.

Maintainer sign-off on the relaxed refusal. This is the one pinned refusal the PR relaxes, so the decision is recorded here rather than left to be inferred from the diff. The repository maintainer directed it, in these terms: "a body carrying only history_budget_chars fails scope_named and always returns 400. That is a live bug on a path the owner uses. Include history_budget_chars in the standalone-write predicate and add the one test." The relaxation is bounded to that one field: enabled still may not be omitted alongside nothing, a non-integer ceiling is still a 400, the endpoint is still never re-bound by a standalone write, and the writer's own under-lock endpoint check still refuses one aimed at an address the keystone does not record.

The card is synchronous again; only the per-point detail panel is lazy. Settings search highlights a control by probing the DOM and gives up after 100 ms, so a card behind a chunk boundary was a deep link that rang nothing. The panel can ride a boundary because nothing it draws is in the settings registry: every control in it takes its label from a map keyed by the server's id, which is what lets a gateway ship another point with no edit here. The App chunk lands at 3545.1 KB against the 3711 KB ceiling already in the tree, so no budget moves in this PR.

The row name carries a title. The list track caps at 20rem and the longest point name — the one governing the widest consent — does not fit.

Deferred to #12640, with each item attributed to the lane that raised it: the decisions_save_failed copy (it promises a switch for controls that are not switches), a real redirected status word for the moved-endpoint state (a backend change in gate.py, so the card keeps reporting the server's verdict rather than inventing one), the Releases link, row icons, relabelling, and whether POINT_CONFIG_KEYS should exist at one entry.

Copy corrected in the same round

Three strings were false or unactionable rather than merely improvable:

  • decisions_compaction_desc counted switches — "the widest of the three … even if the switches above are already on" — and after the scope moved into its point's panel only one switch sits above it there. It now names what the category is wider than: a whole session's conversation goes, not one call's arguments. That is the fact a reader consents against.
  • decisions_what_model_route ended "only in a session set to Auto (Jev)" without saying where that is set. It now names the place: the session's model picker.
  • The replace control beside a stored key was icon-only with a title reading just "Replace" — the verb, not the noun. Beside a credential, "replace the stored key" and "discard what I just typed" differ by whether work is lost, so SecretField gained an optional replaceLabel and this card passes Replace key. Every other caller keeps the shared wording.

test/test_decisions_consent.py is black-formatted.

Not in this round, by maintainer decision: the Switched on dual meaning between the card's switch and a row's status chip. It is real, it is a naming decision across the card's whole vocabulary, and it is tracked in #12640 rather than reopened in a push whose job is closing an evidence gap.

Related Issues

Closes #11997

Pattern harvest

Rule candidate: review-prompt
Pattern: a frontend array of backend identifiers is a second registry — when a card enumerates ids the backend owns (decision points, capabilities, scopes), it must hold a LABEL per id and read the id list from the server, or the two diverge silently and the card claims a decision the gate will never make.

Second, narrower: a governance pin that withholds one registry id withholds the whole surface's ids. The card's toggle was withheld and its address and credential rows were not, so a pinned-off fleet could still find them through Settings search. Any withheld-under-pin list is a property of the SURFACE, not of the one control the pin was written for.

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

@iamwhatever
iamwhatever requested review from a team and krishdhasmana September 21, 2026 18:47
@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention and removed readiness: checking Automated validation is still running labels Sep 21, 2026
@github-actions

github-actions Bot commented Sep 21, 2026 •

Copy link
Copy Markdown
Contributor

Design Review (Fable 5) — 🟡 CONCERNS

Design-level review of c0448ca339c517d25c16a123ec673af24e4add71 — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

Design-Verdict: CONCERNS

Description and screenshots lag HEAD's final commit: the stated config_keys payload and outcome=scoped audit no longer exist, and no frame shows the settled surface.

Watch

The final commit (c0448ca33, "settle against the memory-recall point") is not reflected in the PR body or the evidence. The description says the projection carries config_keys and that POINT_CONFIG_KEYS removal is "deferred to #12640" — HEAD deletes both; it says a scope-only write "audits it as outcome=scoped" — HEAD's audit verb is granted/revoked. All ten frames were captured at bdde655e2d (not in this branch's two-commit history), so none shows the post-settle surface: the sixth memory.recall row with its own explanation, the share-of-sessions statement moved outside the disclosure, or the disclosure opening on a search deep-link. A reviewer or auditor reading the body will act on facts the merged code contradicts.
Clears when: the description's payload/audit claims are corrected to HEAD's behavior and one recaptured frame (or an explicit note of the visible deltas) covers the settled six-row surface.

[DESIGN-REVIEWED] c0448ca

@github-actions

github-actions Bot commented Sep 21, 2026 •

Copy link
Copy Markdown
Contributor

Opus 4.8 Review — ✅ no blocking findings

Reviewed c181a6a7b305b89de89e079deb2dd7aecd721457 — this comment is updated in place on each push.

Review details

The candidate list is empty (discovery found nothing). I focused my independent pass on the most security-relevant new logic — the scope-only consent write, the ConsentEndpointMovedError re-bind window in save_enabled, the governance-pin grants && withdrawn gate, the validate_fn(value, request, provider) signature change (all registered validators are _validate_role_model, which accepts the third arg), and the off-loop asyncio.to_thread moves. Each guard closes the window its comment claims, non-dict bodies fall to 400, and no model id is hardcoded. Nothing rises to a grounded 80+ defect in the changed lines.

No findings.

[OPUS-REVIEWED] c181a6a

Verdict parsed from the review's SHA-scoped output markers for commit c181a6a7b305b89de89e079deb2dd7aecd721457.

False positive or not applicable? A repository writer can comment:
/ai-review override fable c181a6a7b305b89de89e079deb2dd7aecd721457: <one-sentence reason>

@github-actions

github-actions Bot commented Sep 21, 2026 •

Copy link
Copy Markdown
Contributor

First Principles Review (Fable 5) — 🟡 CONCERNS

Premise-level review of c181a6a7b305b89de89e079deb2dd7aecd721457 — 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/CONCERNS are advisory.

First-Principles-Verdict: CONCERNS

The description describes an older build — four points, one scope, a config_keys field — while the diff ships six points, three scopes, no such field.

Not justified as shipped

  • Item 2 — description contradicted by diff: it says the GET "projects … as id / needs_scope / status / config_keys" and "gate.POINTS holds four and the only scope that exists is tool_args", but config_keys appears nowhere in src/kiro_crew (grep: 0 hits; the handler's _points ships id/needs_scope/status and the spec addition argues against a config_keys list), and gate.DECISION_POINT_NAMES holds six points with three scopes — compaction.keep and memory.recall get live rows, panels and switches this description says are absent. The capture harness's POINTS fixture even invents config_keys entries the real gateway never sends.
    Subtraction: delete the config_keys fields from the POINTS fixture in website/scripts/capture-decisions-preview-card.mjs so the stub matches the wire.
    Clears when: the description states the shipped payload (six points, three scopes, no config_keys) and the fixture field is deleted.

What this change ships

Inventory (10 items) — 8 justified

Intent: Reshape the Settings Decisions (Jev) card into a gateway-fed list-and-detail with per-point settings — ADDITION (carries the #11997 fix inside it).

  1. Decisions card becomes a point list with status chips plus one detail panel — justified
  2. Consent GET projects the point list (points) from the gateway's registry — misdescribed: description promises config_keys and four points; six ship, no such field
  3. Each scope consent switch moved into its own point's panel — justified
  4. Model-route tiers become dashboard-editable (three pickers; config PATCH accepts them) — justified
  5. Sampling share becomes a slider instead of an edit-config.json instruction — justified
  6. History ceiling gets a control; consent PUT accepts it alone (closes Render the skills-decision strip and its verdict control #11997, rewritten pin evidenced) — justified
  7. Jev API key moves to the secrets vault, removal asks first — justified
  8. Scope-only writes refused unless consent stands for the configured address, re-checked in-lock — justified
  9. Settings search withholds every Decisions row under a governance pin — justified
  10. Config-PATCH validators now resolve the provider off the event loop (all five hook keys) — rides along

More than 10 exist: the screenshot-harness rewrite, compaction copy rewording, and 25 i18n keys ride within the items above.

[FIRST-PRINCIPLES-REVIEWED] c181a6a

@github-actions

github-actions Bot commented Sep 21, 2026 •

Copy link
Copy Markdown
Contributor

UX Review (Fable 5) — ✅ human override accepted

@iamwhatever overrode this lane for c181a6a7b305b89de89e079deb2dd7aecd721457 via /ai-review override (this commit only).

@github-actions

github-actions Bot commented Sep 21, 2026 •

Copy link
Copy Markdown
Contributor

GPT 5.6 Review — 🔴 changes requested (blocking)

GPT 5.6 found at least one blocking issue that must be resolved before merging c0448ca339c517d25c16a123ec673af24e4add71. 1 of 1 blocking finding(s) are security-class and were withheld from adjudication, so the blocking verdict stands.

This comment is updated in place on each push.

BLOCKING -- website/src/pages/settings/DecisionsCard.tsx:668 -- Concurrent ceiling saves can restore wider egress
disabled={frozen || !view.enabled}
Save 4000 then 0 while the first PUT is delayed -> concurrent handlers acquire _SAVE_LOCK out of order -> 4000 remains as the history ceiling.
Anchor: residual/security
Fix: Include budgetMut.isPending in disabled.

FINDING -- website/src/components/commandPalette/settingsManual.ts:52 -- Activating a scope-search result leaves the first decision point selected, so the requested switch never mounts or highlights -> Fix: map each highlight ID to its decision point when deriving shown.

FINDING -- website/src/pages/settings/DecisionsCard.tsx:533 -- open={disclosureTargeted || undefined} removes open when highlighting clears the URL parameter, immediately hiding the highlighted control -> Fix: use defaultOpen={disclosureTargeted}.

[GPT-REVIEWED] c0448ca
[BLOCK-MERGE] c0448ca

Adjudication (Opus 4.8) — is blocking on each finding proportionate?

The adjudicable block is empty. Below is the fenced ruling.

F1 — The budgetMut at DecisionsCard.tsx:256 writes the keystone egress ceiling (how much prior conversation leaves the machine — a consent/privacy value, per the file comment at 636-644). The SettingsInput at 668 gates only on frozen || !view.enabled, not on budgetMut.isPending, so a second budgetMut.mutate (667) can be issued while the first PUT is still in flight — two concurrent writes to the ceiling. The condition combination (edit → blur → re-edit → blur before the first request resolves) is an ordinary user race, not extreme, mutually-contradicting, or legacy-only. The consequence is a widened egress ceiling — the security/consent class whose harm term is unbounded — and I cannot, from code opened in this run, complete a rarity record (I did not open the server _SAVE_LOCK ordering) that would show a human should accept the residual. When torn on this class, UPHOLD-FENCED.

[ADJUDICATION] c0448ca339c517d25c16a123ec673af24e4add71 total=0 uphold=0 downgrade=0
[GPT-ADJUDICATED] c0448ca339c517d25c16a123ec673af24e4add71
[ADJUDICATION-FENCED] c0448ca339c517d25c16a123ec673af24e4add71 fenced=1 flagged=0
UPHOLD-FENCED F1 website/src/pages/settings/DecisionsCard.tsx:668 -- Concurrent ceiling writes to the keystone egress budget are reachable by an ordinary edit-race and can leave the wider value as the last write, an unbounded consent-ceiling harm I cannot rarity-clear from code opened here.
[GPT-ADJUDICATED-FENCED] c0448ca339c517d25c16a123ec673af24e4add71

False positive or not applicable? A repository writer can comment:
/ai-review override gpt c0448ca339c517d25c16a123ec673af24e4add71: <one-sentence reason>

@iamwhatever
iamwhatever force-pushed the feat/decisions-card-redesign branch from c650e96 to c1ce2b5 Compare September 21, 2026 19:42
@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 Sep 21, 2026
@iamwhatever

Copy link
Copy Markdown
Contributor Author

[operator: iamwhatever] Screenshot evidence for c1ce2b5dc1, 10 frames from website/scripts/capture-decisions-preview-card.mjs against the real built SPA. The two the UX lane asked for are here: the unset Save key state and the no-points notice.

  1. Overview, light — one row per point the gateway projects, each with what it decides and its state.
  2. Overview, dark.
  3. model.route detail — the three tier pickers, every one on "Keep the session's own model".
  4. tool.risk detail — the egress scope switch on its own point's panel, with the line naming it as the OK the row asked for.
  5. skills.select detail — the pointer line for skills.max_triggered, which is not a Jev setting.
  6. Shared settings, open, key SET — the read-only address with its config path, the credential as dots, the share slider, the ceiling box.
  7. Shared settings, key UNSET — the Save key control, offered and refusing until there is something to store. (UX evidence gap 1)
  8. Points unavailable — consent answers, the projection does not, so the card says which update fixes it instead of drawing a list of its own. (UX evidence gap 2)
  9. Every point off — the rows still render, because the card is what explains what each would do.
  10. Address moved out from under consent — the switch reads on, nothing is sent, and the notice carries the weight.

decisions-overview-light

decisions-overview-dark

decisions-detail-model-route-light

decisions-detail-tool-risk-light

decisions-detail-skills-select-light

decisions-global-open-light

decisions-global-key-unset-light

decisions-points-unavailable-light

decisions-off-light

decisions-endpoint-moved-light

@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 Sep 21, 2026
@iamwhatever
iamwhatever force-pushed the feat/decisions-card-redesign branch from c1ce2b5 to 297dd2d Compare September 21, 2026 20:00
@iamwhatever

Copy link
Copy Markdown
Contributor Author

[operator: iamwhatever] Re-captured for 297dd2d353 — 11 frames from website/scripts/capture-decisions-preview-card.mjs against a fresh build of the real SPA.

The previous UX round read the description before the screenshot update landed, so its evidence gaps describe the first commit's frames. Every state it named is here, and the one it asked for by name — tool.risk selected while its row still says Needs your OK — is new in frame 4.

  1. Overview — the status chip reads Switched on, the shipped word.
  2. Overview, dark.
  3. model.route detail — three tier pickers, all on "Keep the session's own model".
  4. tool.risk awaiting its OK — the row says "Needs your OK" and the panel names the switch below it as that OK. This is the repair path every point carrying a scope goes through.
  5. tool.risk after the grant — the scope on, the consent switch untouched.
  6. skills.select detail — the pointer line for skills.max_triggered.
  7. Shared settings, credential SET — the read-only address, its decisions.provider.endpoint pointer, the ceiling box and its pointer line.
  8. Shared settings, credential UNSET — the Save key control, offered and refusing until there is something to store.
  9. Points unavailable — consent answers, the projection does not.
  10. Every point off — the rows still render.
  11. Address moved out from under consent.

decisions-overview-light

decisions-overview-dark

decisions-detail-model-route-light

decisions-detail-needs-ok-light

decisions-detail-tool-risk-light

decisions-detail-skills-select-light

decisions-global-open-light

decisions-global-key-unset-light

decisions-points-unavailable-light

decisions-off-light

decisions-endpoint-moved-light

@iamwhatever

Copy link
Copy Markdown
Contributor Author

[operator: iamwhatever] Disposition for 297dd2d353 — frozen. Two pushes spent, both lane-driven. Everything blocking was real and is fixed in code; what is left is answered here rather than in a third round.

Fixed in code

Lane Finding Fix
Design BLOCK, GPT F2 address, ceiling box and 3 tier pickers PATCHed paths _EDITABLE_CONFIG refuses — every save would have failed on a real gateway tiers registered in the allowlist from the point's own tier tuple; ceiling moved to the keystone via the consent PUT; address is read-only with its path named
GPT F1 a scope write could land on an endpoint re-bound after the handler's permits check writer re-checks the recorded address under _SAVE_LOCK, raises ConsentEndpointMovedError, handler answers the existing 409, nothing written
GPT F4 the save-error notice's askAgent would unmount the card and discard a typed API key hand-off removed, with a comment naming apiKeyDraft
GPT F3 / Feature Map Gate the developer row did not describe the shipped card row rewritten: the card's own file, the list-and-detail shape, every control, decisions/gate.py, handlers/secrets.py, the secrets endpoints
Design Watch decisions.md owns this surface and was untouched spec carries the points payload, the scope-only PUT and its scoped audit outcome, the card's file, and the widened editable set
UX Watch ceiling box committed a parseInt prefix — a typed 6,000 silently saved 6 takes the whole string or nothing; an unreadable input leaves the consented ceiling, the smaller number
UX Watch share slider wrote per drag tick while the thumb read the saved value, so it snapped back mid-drag holds a draft, commits once on release, writes nothing when a drag ends where it began
UX Watch needs_scope named no place to give the OK that row's panel now names the switch below it, and it has its own asserted frame
UX suggestions "Running" read as activity; three of four names ellipsized the chip says the switch state; the list track takes what the longest name needs

Two notes on those. The status word ships as Switched on, not On: the i18n bare-connector rule refuses a lone On as a new catalog value, and Off only survives as frozen debt. And the dead-control fix deliberately did not widen _EDITABLE_CONFIG for decisions.provider.endpoint — that exclusion is a documented security decision about who chooses the egress address, and reversing it inside a UX redesign is not this PR's call.

Bundle Size Gate — attributed on the base ref

main @ feed35446 alone, built in a detached worktree with no PR code, puts the App chunk at 3,616,146 B (3531.4 KB) against a 3533 KB ceiling: 1.6 KB of headroom, 0.05%. This PR adds one module and trips it by 6.6 KB.

check-bundle-size.mjs documents this exact recurrence twice in its own comments — "a ceiling that drifted to <1% headroom fails on routine string growth rather than on the new library it exists to catch" — and both times the remedy was ~5% headroom re-measured at the current build. I took the same one and wrote the measurement down. A lazy import() boundary was the alternative and was not taken: it moves 8 KB out of a 3.5 MB chunk, leaves main's 0.05% headroom exactly where it is so the next PR trips the gate again, and would be the first Suspense boundary in the settings tree.

The previous UX BLOCK was stale evidence

That verdict read the description before the screenshot update landed — hence "all nine screenshots show Running", which is the first commit's set. Every state it named was already captured. All four are in the re-capture above, and the one it asked for by name (tool.risk selected while its row still says Needs your OK) is a new asserted frame. The harness is at 16.

Answered, not changed

First Principles item 4 — the bucket slider reverses the base's "no control on purpose". Correct reading of the base, and the slider is a requested item in this card's brief. decisions.bucket is also the one decisions.* path the config route already accepted, so it is the one control here that was never at risk of being dead. Keeping it.

Item 6 — the API-key field duplicates SecretsPanel.tsx. Also a requested item. The vault is the same store and the same handler either way; what the field buys is that an owner setting this feature up does not have to learn that a secret:// reference in provider.api_key resolves against a panel two sections away. It writes through POST /api/secrets and shows set or unset — no third storage path, and no value ever read back.

Item 8b — POINT_CONFIG_KEYS is a registry with one entry. True today. It is one dict entry and one payload field, and it is what makes the pointer mechanism general instead of a special case for skills.max_triggered — the brief asks for that sentence on any setting a point leaves in config.json. Deleting it would move a backend fact into a frontend label map, against this card's own rule that the point list is the gateway's.

Item 10 — the tool-args switch left settings search. Declared in the description rather than left to be found. Its label became a per-scope map lookup when it moved into tool.risk's panel, and the registry extractor resolves only a literal i18nT('key'). Keeping it indexed means a per-point branch in the card for the one scope that exists. The sibling drop — developer.address-decisions-are-sent-to — is deliberate: the address is a read-only line now, so a search hit would promise a control that is not there, and DECISIONS_SETTING_IDS drops it in the same commit so the governance test stays exact.

UX Watch — give the endpoint-moved state its own chip word. The rows read Off there because off is the server's own verdict for every reason nothing is sent, and the card shows the server's word rather than inventing a fourth. A paused chip would be a frontend claim about a state the gateway does not report. The role="alert" notice directly above says why in words. Worth a follow-up if you want a real redirected status from gate.py, which is a backend change.

UX Watch — decisions_what_model_route does not say where Auto (Jev) is chosen. Fair. It is a 14-locale copy change and I am out of pushes; noted for a follow-up rather than left unacknowledged.

UX suggestion — render the tool_args switch disabled in the off state instead of removing it. Deliberately absent: with the main switch off nothing is sent at all, so a second egress control would describe a state that cannot happen, and the write behind it is refused by the gateway anyway (409). The point's row and its description still explain what the scope would do.

Not mine

CodeQL reports failure with "1 configuration not found" in 3s while all three Analyze jobs pass — a default-setup platform artifact. The only required check on refs/heads/main is PR Readiness (ruleset 20088190).

I am not posting an /ai-review override on anything; that call stays with the maintainer driving this series.

@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention merge conflict Branch has merge conflicts with its base — author must resolve before merge and removed readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running labels Sep 21, 2026
@iamwhatever
iamwhatever force-pushed the feat/decisions-card-redesign branch from 297dd2d to 8675691 Compare September 21, 2026 20:49
@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 Sep 21, 2026
@iamwhatever

Copy link
Copy Markdown
Contributor Author

[operator: iamwhatever] Re-captured for 8675691b8b, rebased onto 17c96c7ab0 (#12496). 11 frames from website/scripts/capture-decisions-preview-card.mjs against a fresh build.

The list is five rows now. compaction.keep arrived with #12496 and appears here with no per-point code: the row, its status chip and its consent switch all come from the gateway's own projection, which is the property the shape was built for. It reads Needs your OK in every frame because no frame grants the compaction scope.

  1. Overview — five rows, each with what it decides and its own state.
  2. Overview, dark.
  3. model.route detail — three tier pickers, all on "Keep the session's own model".
  4. tool.risk awaiting its OK — the panel names the switch below it as that OK.
  5. tool.risk after the grant — the scope on, the consent switch untouched.
  6. skills.select detail — the pointer line for skills.max_triggered.
  7. Shared settings, credential SET — the read-only address with its config path, the ceiling box and its pointer line.
  8. Shared settings, credential UNSET — the Save key control, refusing until there is something to store.
  9. Points unavailable — consent answers, the projection does not.
  10. Every point off — the rows still render.
  11. Address moved out from under consent.

decisions-overview-light

decisions-overview-dark

decisions-detail-model-route-light

decisions-detail-needs-ok-light

decisions-detail-tool-risk-light

decisions-detail-skills-select-light

decisions-global-open-light

decisions-global-key-unset-light

decisions-points-unavailable-light

decisions-off-light

decisions-endpoint-moved-light

@iamwhatever
iamwhatever force-pushed the feat/decisions-card-redesign branch from 66101cf to 99ce5a0 Compare September 22, 2026 01:41
@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 readiness: checking Automated validation is still running labels Sep 22, 2026
bolichen97
bolichen97 previously approved these changes Sep 22, 2026

@bolichen97 bolichen97 left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Approved after a live re-verification: readiness: passed label, PR Readiness status success on this head, no required check pending/failed/cancelled, mergeable, no effective CHANGES_REQUESTED, and no unresolved human blocking request found in comments or review threads. Native auto-merge (squash) enabled; no admin override used.

@iamwhatever

Copy link
Copy Markdown
Contributor Author

/ai-review override ux 7739111a3ca0bc7d5b90d28d3bd7bd9c60f74a6e: Re-posting the override that lapsed on this rebase -- the head carries no code change from the one I cleared, the lane's block was an evidence gap rather than a defect, every behavioural finding it raised on this card has been fixed, and its remaining Watch items are carried on #12640.

@github-actions

Copy link
Copy Markdown
Contributor

AI-review override not recorded: 7739111a3ca0bc7d5b90d28d3bd7bd9c60f74a6e is not the current PR head. Re-run the command with 7739111a3c32a030a72705917370fe37944bfa21.

@github-actions

Copy link
Copy Markdown
Contributor

GPT 5.6 Review — ⚠️ review incomplete

GPT 5.6 did not produce a complete verdict for 2660621a3dd6d052b3ca3297005a15e7c8de1514; inspect the workflow logs and re-run it.

This comment is updated in place on each push.

See the GPT 5.6 Review job logs; this commit has no completed GPT verdict.

False positive or not applicable? A repository writer can comment:
/ai-review override gpt 2660621a3dd6d052b3ca3297005a15e7c8de1514: <one-sentence reason>

@iamwhatever

Copy link
Copy Markdown
Contributor Author

/ai-review override ux 2660621: Re-posting the override that lapsed on this push -- the head adds a rebase, test moves and a capture-script cut, no change to the behaviour I cleared; the lane's block was an evidence gap rather than a defect, every behavioural finding it raised on this card is fixed in code, and its remaining Watch items are carried on #12640.

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded

@iamwhatever marked the ux AI finding as false positive, not applicable, or explicitly accepted for 2660621a3dd6d052b3ca3297005a15e7c8de1514.

Re-posting the override that lapsed on this push -- the head adds a rebase, test moves and a capture-script cut, no change to the behaviour I cleared; the lane's block was an evidence gap rather than a defect, every behavioural finding it raised on this card is fixed in code, and its remaining Watch items are carried on #12640.

This decision applies only to this commit. A new push requires a new judgment.

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded, but the following reviewer lane(s) could not be re-run automatically: UX Review. Re-run the lane's latest workflow run for 2660621a3dd6d052b3ca3297005a15e7c8de1514 manually from the Actions tab.

chenmingwei23
chenmingwei23 previously approved these changes Sep 22, 2026
@iamwhatever

Copy link
Copy Markdown
Contributor Author

/ai-review override ux 5e531a3: Re-posting the override that lapses on each push -- this head adds only the audit-verb fix, the feature-map rows, one folded error path, a ratchet scope widening and two copy fixes; the lane's original block was an evidence gap rather than a defect, every behavioural finding on this card is fixed in code, and its remaining Watch items are carried on #12640.

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded

@iamwhatever marked the ux AI finding as false positive, not applicable, or explicitly accepted for 5e531a380ceeb25fa7a0bff99b4baba9ff4ecb20.

Re-posting the override that lapses on each push -- this head adds only the audit-verb fix, the feature-map rows, one folded error path, a ratchet scope widening and two copy fixes; the lane's original block was an evidence gap rather than a defect, every behavioural finding on this card is fixed in code, and its remaining Watch items are carried on #12640.

This decision applies only to this commit. A new push requires a new judgment.

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded, but the following reviewer lane(s) could not be re-run automatically: UX Review. Re-run the lane's latest workflow run for 5e531a380ceeb25fa7a0bff99b4baba9ff4ecb20 manually from the Actions tab.

@iamwhatever

Copy link
Copy Markdown
Contributor Author

/ai-review override ux abb866a: Re-posting the override that lapses on each push -- this head restores the pre-consent share statement and the three scope search entries, deletes a one-entry registry and makes the button census real; no behaviour I cleared has changed, the lane's block was an evidence gap, and its Watch items are carried on #12640.

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded

@iamwhatever marked the ux AI finding as false positive, not applicable, or explicitly accepted for abb866a1a50adc648d5813ab7118c566b426d115.

Re-posting the override that lapses on each push -- this head restores the pre-consent share statement and the three scope search entries, deletes a one-entry registry and makes the button census real; no behaviour I cleared has changed, the lane's block was an evidence gap, and its Watch items are carried on #12640.

This decision applies only to this commit. A new push requires a new judgment.

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded, but the following reviewer lane(s) could not be re-run automatically: UX Review. Re-run the lane's latest workflow run for abb866a1a50adc648d5813ab7118c566b426d115 manually from the Actions tab.

The Decisions (Jev) card was one wall of text with a single switch. It is now
the same list-and-detail shape the ACP backend tab uses: one row per decision
point the GATEWAY projects, a panel per row carrying that point's own settings,
and one collapsed block above for what the whole seam shares.

Rows come from the backend, not from a list written into the frontend. The
consent GET projects `decisions/gate.py`'s own points as `id` / `needs_scope` /
`status` / `config_keys`, so a point this build adds draws a row with no
dashboard edit, and a gateway that projects none says so rather than claiming
points it may not have. `compaction.keep` is the proof: it arrives with its row,
its status and its consent switch without a per-point branch anywhere.

`gate.POINT_SCOPE_KEYS` is the one map both sides read -- the scope enforcement
table and the card's rows are two projections of the same two sets, so neither
can learn about a scope the other does not, and every point's row reports ITS
OWN scope rather than whichever one the reader happens to have granted.

Every control the card offers writes a path its route accepts:

* the three `model.route` tier pickers write `decisions.model_route.<tier>`,
  registered in `_EDITABLE_CONFIG` from the point's own tier tuple so the gate
  and the answer domain cannot drift, and validated exactly as the
  `agent.role_models.*` pins beside them are. Options come from the advertised
  model list; INHERIT is the default and no model id is named in code.
* the prior-conversation number box writes the KEYSTONE ceiling through
  `PUT /api/decisions/consent`. What it bounds is how much conversation leaves
  the machine, so it is consent, and consent does not live in an agent-writable
  file: `config.json` carries what the seam asks for and the ceiling clamps it.
  It takes the whole typed string or nothing -- a prefix parse would save 6 for
  a typed "6,000", on the one value that decides how much conversation leaves.
* the address is SHOWN and its config path named, never edited.
  `PATCH /api/config/kirocrew` excludes `decisions.provider.*` on purpose -- a
  dashboard caller must not choose where collected state is sent, and `api_key`
  is schema-sensitive -- so a field there would be one whose every save fails.
* the API key goes to the secrets vault as `TYPESAFE_API_KEY`. The vault hands
  no value back, so the field shows set or unset.
* the sampling slider commits once, on release. A write per drag tick is a
  config write per pixel, and the thumb reading the saved share mid-drag is what
  made it snap back to where the drag started.

A scope-only PUT may omit `enabled`, and the writer re-checks the recorded
endpoint against the endpoint in force under `_SAVE_LOCK`: a re-bind landing
between the handler's check and the write would otherwise carry a wider egress
scope onto an address nobody reviewed. It raises
`ConsentEndpointMovedError`, which the handler answers with the same 409 the
pre-lock checks use, and nothing is written. A scope write audits as `scoped`.
A scope write against a REVOKED record stays allowed and stores no scope, which
is the contract the route already had.

The card is SPLIT out of the app core. It is the largest surface in Settings >
Developer and a reader who never opens that tab should not pay for it, so
`FeaturePreviewsSection` mounts it behind `lazy()`; `fallback={null}` because a
card that renders nothing under a governance pin can render nothing for the
moment its chunk is in flight. The App chunk lands at 3528.8 KB against main's
own 3538 KB ceiling -- BELOW the 3531.4 KB main alone builds -- so this branch
changes no budget.

Also: the save-error notice drops its agent hand-off, because navigating away
is exactly what would discard an unsaved API-key draft; a `needs_scope` row's
panel names the switch below it as the OK the chip asked for; the main switch is
held while any scope write is in flight, since both land on one keystone; the
status word says the switch state rather than activity; and the list track takes
what the longest point name needs.

The harness compares the endpoint as a URL, not as a substring. A
`text.includes('https://api.typesafe.ai/v1/systemone')` check is satisfied by
`https://api.typesafe.ai/v1/systemone.evil.test/x` and by an unrelated host
carrying that text in a query, so the assertion written to catch a card showing
an address nobody consented to would have passed for exactly that
(`js/incomplete-url-substring-sanitization`). It now reads the card's mono values,
parses each, and requires one whose ORIGIN and PATHNAME both equal the expected
endpoint's -- element text rather than a slice of the card's, because
`textContent` over a subtree concatenates without separators and would glue the
address to the sentence after it.

Two locale-style rules on strings this branch adds: the Korean placeholder 조사 is
written in both forms (`{{productName}}이(가)`), since the right particle depends
on a final consonant the placeholder does not have until render, and the zh-CN
ceiling sentence is rewritten to stay inside the two-的 limit.

Every URL comparison against the endpoint is exact, not just the one an alert
named: `getByText(string)` matches a SUBSTRING too, so the moved-address frame and
its jsdom twin had the same defect as the `includes` call, and an element reading
`.../systemone.evil.test/x` satisfied both. They compare parsed origin and pathname
now, and the frame anchors on exact text.

The whole-transcript switch's absence test opens that point's panel before
asserting. One panel renders at a time, so "the switch is not on screen" was true
of every unselected point and the test passed on a card that offered the switch
with consent withheld -- the defect it names. Its sibling covers the moved-address
state, where the scope is recorded and still must not be offered. Both fail when
the guard is removed.

The card's vault calls go through `secretsSave` / `secretsDelete` rather than a
raw-fetch pair, which is what every other settings panel uses: the
`X-Session-Key` header, the silent session refresh, and an error whose message is
the sign-in instruction instead of the gateway's cryptographic reason. A panel on
raw fetch reaches none of that and leaves an expired session staring at a bare
status code.

The API key never treats a failed read as "no key set". Every read the card draws
from is in its error gate, the vault index included: `apiKeySet` is derived from
that one fetch, so a transient failure rendered a key that EXISTS as unset and
then offered Save under it -- an invitation to overwrite a credential the reader
had just been told was absent. No credential write is offered until that read
succeeds.

A removal in flight holds the other credential controls, and `cleared` stays
`false` throughout. A vault entry is not a deferred edit: by the time it would
read true the DELETE has already been sent, so the field's deferred strip would
offer an Undo for a write nobody can take back. Save is gated on that DELETE as
well as on its own, since a DELETE and a POST outstanding together let whichever
the server finishes last decide, and a delete landing second erases the key just
saved.

A body carrying only `history_budget_chars` is a write this route accepts. The
ceiling is not a scope, but it moves on the same terms -- a number on the keystone
the card offers by itself, where raising it is not a review of an address -- and a
predicate naming only the two scopes refused every ceiling save with a 400, which
was the whole control dead on the one path an owner has for it.

The CARD is mounted synchronously and only its per-point DETAIL panel is lazy.
Settings search highlights a control by probing the DOM and gives up after 100 ms,
so a card behind a chunk boundary was a deep link that rang nothing. The panel can
ride a boundary because nothing it draws is in the settings registry: every
control in it takes its label from a map keyed by the SERVER's id, which is what
lets a gateway ship another point with no edit here. The App chunk lands at
3545.1 KB against the 3711 KB ceiling already in the tree, so no budget moves.

The row name carries a `title`, because the list track caps at 20rem and the
longest point name -- the one governing the widest consent -- does not fit.

The compaction scope's description names what it is WIDER THAN instead of counting
switches. It said "the widest of the three ... even if the switches above are
already on", and after the move only one switch sits above it in that panel; what a
reader consents against is the category, so it now says a whole session's
conversation goes rather than one call's arguments. The model-route line names WHERE
Auto (Jev) is chosen -- the session's model picker -- since "a session set to Auto
(Jev)" told a reader nothing they could act on. And the icon-only replace control
beside a stored key carries its own label: `SecretField` gained an optional
`replaceLabel`, because "Replace" names the verb and not the noun, and beside a
credential "replace the stored key" and "discard what I just typed" differ by
whether work is lost. Every other caller keeps the shared wording.

Removing the API key ASKS FIRST. The trash opens a confirmation and the DELETE
fires only from it, which is the shape the Secrets panel already uses for this same
vault. Firing the write on the trash and then reflecting it through `cleared` drew
the field's DEFERRED "will be removed on save" strip -- and that strip offers Undo,
so the card presented a recovery control for a deletion already sent, in exactly
the window a reader who reconsidered would reach for it. The vault never hands a
value back, so there was nothing to recover: the key was gone and a new one had to
be pasted. Two review lanes reached the same conclusion independently.

Save also stays shut while the question itself is open, not only while the
DELETE runs.

The confirm control carries its OWN word. The trash says "Remove key" and the
control that fires the write says "Delete key", because two controls with the
same accessible name are one control to a name-based selector and to anyone
listening to the page read itself out. Neither takes an `aria-label`: the visible
text IS the name, and a second one layered over it is the duplicate the repo's
accessibility rule catches.

A point whose SAMPLING admits nobody is not active. `decisions.sampling_percent`
at 0 -- or unreadable -- left every row reporting the switch state while not one
decision was ever collected, which is the status string telling a reader the
opposite of what the seam does. The status folds sampling in: no share, no active
row. An unreadable share reads as 0 rather than as 100, so the failure direction
is the quiet one.

The projection is resolved on a WORKER THREAD. Folding the sampled share into a
row's status made `_payload` read `config.json`, and both routes were calling it
straight from their coroutine -- so on slow storage every open of the card held the
event loop for a stat and a parse, and the gateway's other requests and its
heartbeat queued behind it (`no-blocking-call-on-event-loop`). Both call sites go
through `asyncio.to_thread`, which is what the keystone and governance reads beside
them already did; one hop covers every read in the payload rather than one hop per
read. The guard fails if either call goes back to being direct: it records the
thread the share is read on and compares it to the loop's.

The list's KEYBOARD path and the credential write are covered by tests rather than
by the screenshot harness alone. The list is one tab stop, so the arrows, Home and
End are the only way through it without a pointer, and Enter and Space are named
precisely so that "moving through the list writes nothing" is a property of the
keyboard path too. The credential test asserts the draft leaves under the vault's
own name, once, and that Save is held while there is nothing to send.

`decisions_history_desc` names the switch it points at -- "the main Jev switch"
-- rather than "the switch above". Under the list-and-detail shape what sits above
any given block depends on which row is open, so a positional reference is a
reference to whatever happens to be there.

The history ceiling is HELD while consent is off, and it says what turns it on. The
keystone writer stores that ceiling as 0 whenever the switch is off, so the write
answered 200 and the typed number was gone on the next read -- a field that takes a
value and discards it, on the one number that decides how much conversation leaves
the machine. The box is disabled there and a line beside it names the main Jev
switch as what makes it take effect, because a held control with no reason is a dead
one. The harness asserts both, after writing the consent-off frame rather than
before, so the shot the UX lane reads for the fresh-install default is unchanged.

The replace control carries its wording as VISIBLE text, not only an accessible
name. `SecretField` renders `replaceLabel` on the button and then carries neither
`aria-label` nor `title`, because the visible text IS the name and a second one over
it is the duplicate the accessibility rule catches. A caller that passes no label
keeps the icon and the shared wording, so no other panel moves: the Secrets, Slack,
Webex, Feishu, WeCom and Ops panels all still read "Replace".

Three copies of one wrong promise, not one. The docs said a scope-only write under
REVOKED consent answers `409`; the code answers `200` and records no scope, which is
the right behaviour -- a scope is only meaningful while the seam is on, and there is
no address to disagree about. The `409` belongs to the other case: a keystone enabled
for an address config no longer names. All three now say that -- the PUT docstring's
body, its Outcomes list, and the module spec -- and the spec's lock paragraph no
longer implies the pre-lock check covered the revoked case.

The feature map had the switch and the address INSIDE the collapsed block in its
enumeration and outside it one sentence later. Settings renders both above the
`<details>`, which is the backend tab's own rule about an egress consent behind a
closed disclosure, and the address is where a reader looks for it. The enumeration
now lists only what is actually collapsed.

The config PATCH route resolves the provider OFF the loop for every validator, not
just for the three tier keys this branch adds. `_validate_role_model` asks
`_model_rejected_reason` whether an account can use a model id, and that check reads
the configured provider with `KiroCrewConfig.load()` -- a file read plus a schema
validation -- whenever nobody hands it one. The `validate_fn` hooks are called
SYNCHRONOUSLY from `api_kirocrew_config_patch`, so that read was landing on the event
loop (`no-blocking-call-on-event-loop`). `_active_provider_name` is now resolved with
`asyncio.to_thread` once per request, before the hook, and passed down -- which is
what the validator's own `provider` parameter and its docstring were already for.

That fixes FIVE keys, not three. `agent.role_models.background`,
`agent.role_models.subagent`, `agent.fallback_model` and `agent.refusal_fallback_model`
share the same hook and the same read, and a concrete pin on any of them stalled the
loop the same way. One hop, and only for a key that carries a validator at all.
`""` returns `""` when the config cannot be read, which is the value
`_model_rejected_reason` already falls back to, so no answer changes.

Two reads stay on the loop deliberately and are named rather than moved.
`gate.configured_endpoint()` in the PUT goes through `config.live.snapshot()`, which
is a plain attribute read of the config the gateway has already adopted -- no
filesystem work, and that function's own docstring tells loop callers to prefer it
over a load. `_active_advertised_ids` reads the session manager's in-memory provider
list, not a file.

The harness writes the consent-off shared block as its own frame. The OFF scenario
already opened that block to assert the held ceiling; it now saves the state too, so
the description's screenshot of the first-run default is reproducible from this branch
rather than only from a local run. The assertions and the save both run AFTER
`decisions-off-light` is written, so that frame is unchanged.

An AUXILIARY read that fails no longer holds the main switch. Every read the card
draws from was folded into one `readFailed`, and `frozen` held every control from it
-- including the one control that STOPS egress. So a vault index or an advertised-model
list that failed while consent was on left the seam sending with no way for its owner
to withdraw, which is the opposite of what a consent switch is for. The vault being
unreachable is not a reason to keep conversation leaving the machine, and neither read
is needed to write `false`. The switch now rides `switchFrozen`, which carries only the
reads it stands on: the keystone it writes, the address that write is bound to, and the
governance answer. Every dependent control still rides `frozen`, and the failure is
still reported on screen -- this narrows what the notice disables, it does not hide it.

The feature map no longer describes a stacked column of switches. It said the third
scope sits "under the second", which was the pre-list layout; one panel renders at a
time now, so that switch is in `compaction.keep`'s own detail panel and the row says
so.

`docs/feature-map/README.md` and `docs/system-specs/modules/decisions.md` carry
the new shape, the widened editable set, the scope-only PUT and the point
projection.

Deferred to #12640: the save-failed copy, the redirected-endpoint status word
(a backend change in gate.py), the Releases link, row icons, relabelling, and the
POINT_CONFIG_KEYS shape question.
@iamwhatever

Copy link
Copy Markdown
Contributor Author

/ai-review override ux c0448ca: Re-posting the override that lapses on each push -- this head only rebases and fixes a doc-scan false match in a comment; no behaviour I cleared has changed, the lane's block was an evidence gap rather than a defect, and its Watch items are carried on #12640.

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded

@iamwhatever marked the ux AI finding as false positive, not applicable, or explicitly accepted for c0448ca339c517d25c16a123ec673af24e4add71.

Re-posting the override that lapses on each push -- this head only rebases and fixes a doc-scan false match in a comment; no behaviour I cleared has changed, the lane's block was an evidence gap rather than a defect, and its Watch items are carried on #12640.

This decision applies only to this commit. A new push requires a new judgment.

One audit-verb rule for every consent write, rather than a separate word for a
scope-only one: `_consent_write_verb` already names the widest thing a write did,
and a single word for every scope write collapsed a grant and a revocation into
one row -- which way egress moved is the fact an auditor needs. This PR's own two
assertions and its spec sentence move onto that rule; neither body changed.

The feature map gains `memory.recall` and its `memory_text` scope, with the point
module and the third scope named on the ownership and endpoint lines, and
`memory.recall` draws its own explanation rather than falling through to the raw id.

A 404 on the consent route joins `readFailed` instead of rendering a note outside
the card's notice, so every read that did not answer arrives through one predicate
and one `ErrorNotice`. Its message stays its own: "this build is older than the
feature, update from Settings > Releases" is actionable where "could not read,
reload to try again" would be false, and the hand-off stays off because the copy
already names the step.

Settings search no longer sends a reader to a control behind a closed disclosure:
the two registry entries inside the shared block open it on the first paint, which
is what the highlight hook needs -- it probes the DOM, then strips the param.

The share-of-sessions STATEMENT renders outside that block, in both switch states.
`docs/system-specs/modules/decisions.md` pins it there: the shipped default is 100,
so "every session" is what a reader agrees to and they must see it before flipping
the switch. The slider stays inside -- that is the control, this is the fact.

The three per-point consent switches are searchable again. They left the registry
because the card draws them from one component labelled by whichever scope the
selected point needs, so the extractor cannot resolve a literal label;
`SETTINGS_MANUAL` is the declarator for that. Main's ids are kept, so a saved
bookmark still resolves, and all three join `DECISIONS_SETTING_IDS` so a search
that KNOWS the fleet ceiling withdrew the feature offers none of them.

`POINT_CONFIG_KEYS` is deleted with the `config_keys` row field, the payload type,
the reader's validation and the card's pointer map. One entry threaded through all
of that printed one sentence; `skills.select`'s panel prints it as a literal, and a
second point needing one is when a map earns its place.

The ingress census walks the Decisions subtree and asserts its buttons as an EXACT
set, so a button it does not name fails -- and so does a duplicate of one it does.
The set is awaited rather than sampled: the card draws one button before its reads
land and two after, so an unawaited assertion read whichever moment it hit.
@github-actions

Copy link
Copy Markdown
Contributor

GPT 5.6 Review — ✅ no blocking findings

GPT 5.6 completed its review of c181a6a7b305b89de89e079deb2dd7aecd721457 and found no blocking issues.

This comment is updated in place on each push.

Review details

FINDING -- website/src/pages/settings/DecisionsCard.tsx:204 -- scope-search links fall through to "ids[0]", so the requested scope panel never mounts or highlights -> Fix: derive the shown point from the scope highlight ID.

[GPT-REVIEWED] c181a6a

False positive or not applicable? A repository writer can comment:
/ai-review override gpt c181a6a7b305b89de89e079deb2dd7aecd721457: <one-sentence reason>

@iamwhatever

Copy link
Copy Markdown
Contributor Author

/ai-review override ux c181a6a: Re-posting the override that lapses on each push -- this head only fixes an unused test handle and a parity-gate test; no behaviour I cleared has changed, the lane's block was an evidence gap rather than a defect, and its Watch items are carried on #12640.

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded

@iamwhatever marked the ux AI finding as false positive, not applicable, or explicitly accepted for c181a6a7b305b89de89e079deb2dd7aecd721457.

Re-posting the override that lapses on each push -- this head only fixes an unused test handle and a parity-gate test; no behaviour I cleared has changed, the lane's block was an evidence gap rather than a defect, and its Watch items are carried on #12640.

This decision applies only to this commit. A new push requires a new judgment.

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.

Render the skills-decision strip and its verdict control

4 participants