You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit a1b3c64
Browse filesBrowse the repository at this point in the historyBrowse files
Copy file name to clipboardExpand all lines: AGENTS.md
+34-3Lines changed: 34 additions & 3 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -24,6 +24,19 @@ Update corresponding Docs in [CrewCode Docs](/docs/), and [AGENTS.md](/AGENTS.md
24
24
25
25
## Agent Notes
26
26
27
+
Mobile window tabs use the bottom workspace dock's Tabs button and an on-demand
28
+
sheet; never reserve a top tab-strip row on phones. Preserve exact tab selection,
29
+
observed activity, pinned-tab close protection, built-in/plugin new-tab actions,
30
+
and dismiss-on-selection. Desktop retains its tab strip. Mobile overlay offsets
31
+
must not reserve the former 40px strip. See `docs/mobile-responsive-pages.md`.
32
+
33
+
34
+
Claude effort menus include Off, Adaptive, Low, Medium, High, XHigh, and Max.
35
+
Adaptive is a Claude-only UI selection: omit SDK `effort` and `thinking` so
36
+
Claude Code retains its environment/settings/model resolution. Off explicitly
37
+
disables thinking; named levels pass through as SDK effort values. Keep the
38
+
session, crew, desktop/web, and SSH contracts aligned. See `docs/reasoning-effort.md`.
39
+
27
40
The context usage meter must prefer the active provider's reported full context
28
41
window over model-catalog metadata and static limits. Codex app-server usage
29
42
reports `modelContextWindow` as an effective prompt budget; when a known full
@@ -374,8 +387,20 @@ Brain catalogue without replacing genuine identities the Brain already owns, onc
374
387
desktop catalogue authority exists. An explicit Background Brain enable is a new
375
388
owner-observed handoff: before the enable-triggered reload, seed the exact foreground
376
389
desktop chat/session catalogue even when an older authority marker exists. Reconcile
377
-
the current desktop workspace registry first for matching ids/paths and preserve
378
-
genuine Brain-only/browser-created workspaces. Until that marker is present, or while
390
+
the current desktop workspace registry first for matching ids/paths. Preserve only
391
+
Brain-only/browser-created workspaces whose observed creation time is newer than the
392
+
desktop registry; never resurrect older entries missing from the current desktop list.
393
+
Treat explicit Background Brain enable as an audited owner-local grant of each exact
394
+
existing registered local project path, preserving scopes and never granting a broader
395
+
parent; standalone Brain remains explicit-root-only and SSH stays desktop-only. Attached
396
+
Electron must read the exact native registry. Brain/browser catalogue reads may expose
397
+
registered metadata outside current roots as locked rows, but every content or execution
398
+
operation still requires the existing scope, registered-workspace, and live-root gates.
399
+
Never relay device-local project icon data in the workspace catalogue. Keep individual
400
+
encrypted Hub frames within the existing 8 MiB connection-burst ceiling and preflight
401
+
Brain output so an oversized RPC result closes only its logical browser tunnel, not the
402
+
machine relay.
403
+
Until that marker is present, or while
379
404
the Brain still contains transcript-derived recovered rows, Electron reseeds the exact desktop
380
405
names, active selection, tab-key order, and per-tab session order; drop unmatched
381
406
recovered navigation rows while preserving transcript shards, and retain genuine
@@ -452,6 +477,12 @@ The composer PromptPicker has separate Prompts and Skills tabs backed by the sha
452
477
453
478
On phones, Code Editor keeps the code canvas primary and opens its file tree as a dismissible right overlay. Git Sidebar must remain the same stateful surface on desktop and mobile, becoming an off-canvas panel with backdrop/close controls instead of being hidden or squeezing chat/editor content. Changes by turn is a full-screen mobile review: its catalogue stacks above the diff when open, while a targeted changed-file route keeps the catalogue closed. Keep all JS/CSS decisions aligned at `≤768px`; see `docs/mobile-responsive-pages.md`.
454
479
480
+
Mobile/Hub typography must honor the live chat `--mono-size` and CodeMirror
481
+
`--editor-size` settings. Keep the composer at `max(16px, --mono-size)` to prevent
482
+
iOS focus zoom without blocking larger user-selected text, and keep the editor size
483
+
authoritative over CodeMirror's runtime-injected theme. See
484
+
`docs/mobile-responsive-pages.md`.
485
+
455
486
The Code Editor file tree is lazy and must expose dependency, build-output, cache,
456
487
and workspace-local Python environment folders on demand for local and SSH roots.
457
488
Keep `.git` and `.DS_Store` hidden from direct browsing. Do not widen bounded
YuHeard PTY integration must remain bundle-safe. `PtyService` receives the active YuHeard server through an injected accessor and statically imports its shell-wrapper helpers; do not use runtime relative `require('./yuheard-*')` calls from PTY code because electron-vite can move that code into a chunk without emitting the required sibling modules. CLI launch, initial TUI paint, and prompt submission are not completed turns. Codex must use only its exact `approval-requested` and `agent-turn-complete` hook events—never generic PTY idle/BEL heuristics—while output fallback detection remains available for agents without an exact hook. Suppress every YuHeard surface only when the exact completing terminal owns keyboard focus in the focused CrewCode window; a different pane must still alert. See `docs/yuheard.md`.
504
535
505
-
CrewCoder agent profiles are separate from CrewCode execution modes. Show the desktop model-row profile picker only when the installed CrewCoder provider is active; disable it during a running turn, persist the optional session-scoped `crewcoderMode`, omit `--mode` for Configured default, and pass only `general | crewcoder | plugin | extension` to `crewcoder acp --mode`. A concrete profile locks the underlying CrewCode permission policy to Build and disables Ask/Plan/Build/Full on desktop and phone; Configured default re-enables those controls. Never retain a hidden prior Ask, Plan, or Full Access policy under a concrete profile. When the concrete `crewcoder` profile is active, show the separate desktop approval picker and persist `crewcoderApprovalMode`; expose only CrewCoder's `review`, `always`, `never`, `full-access`, and `sandboxed` values, with `review` as the fail-closed default. Treat approval changes as immutable launch authority: disable them during a running turn, drop only the idle bridge, include the value in custody, and native-resume on the next prompt. Never suppresses prompts but continues to block dangerous calls; Sandboxed applies the native sandbox policy where supported; Full access bypasses CrewCoder approval requests and dangerous-command blocking, so label that risk truthfully and never imply CrewCode Build still interposes. Never route Ask/Plan/Build/Full into CrewCoder's `--mode`. The `crewcoder` profile's plan gate is CrewCoder-owned: project `crewcoder_clarify` / `crewcoder_propose_plan` into the activity overlay and send `/approve-plan` as a prompt, never as a tool-permission Allow/Deny. See `docs/crewcoder-provider.md`.
536
+
CrewCoder agent profiles are separate from CrewCode execution modes. Show the desktop model-row profile picker only when the installed CrewCoder provider is active; disable it during a running turn, persist the optional session-scoped `crewcoderMode`, omit `--mode` for Configured default, and pass only `general | crewcoder | plugin | extension` to `crewcoder acp --mode`. A concrete profile locks the underlying CrewCode permission policy to Build and disables Ask/Plan/Build/Full on desktop and phone; Configured default re-enables those controls. Never retain a hidden prior Ask, Plan, or Full Access policy under a concrete profile. When the installed CrewCoder provider is active, show the separate desktop approval picker across general, crewcoder, plugin, extension, and Configured default; persist `crewcoderApprovalMode` independently of profile changes and honor it at launch in every profile; expose only CrewCoder's `review`, `always`, `never`, `full-access`, and `sandboxed` values, with `review` as the fail-closed default. Treat approval changes as immutable launch authority: disable them during a running turn, drop only the idle bridge, include the value in custody, and native-resume on the next prompt. Never suppresses prompts but continues to block dangerous calls; Sandboxed applies the native sandbox policy where supported; Full access bypasses CrewCoder approval requests and dangerous-command blocking, so label that risk truthfully and never imply CrewCode Build still interposes. Never route Ask/Plan/Build/Full into CrewCoder's `--mode`. The `crewcoder` profile's plan gate is CrewCoder-owned: project `crewcoder_clarify` / `crewcoder_propose_plan` into the activity overlay and send `/approve-plan` as a prompt, never as a tool-permission Allow/Deny. See `docs/crewcoder-provider.md`.
506
537
507
538
CrewCoder manual compaction is capability-gated. Expose the bridge `compact()` path only after observing exact `initialize._meta["crewcoder/sessionCompact"].method === "session/compact"`; older CrewCoder versions retain summary-reset. Call the advertised method only while idle and keep CrewCoder's durable session id and ACP child after success. Use the returned summary to replace CrewCode's replay shard and add the visible compact-summary card while retaining the full rich display transcript. Treat `_crewcoder/compaction_update` as authoritative, never infer a duplicate from usage, never fabricate a turn for idle progress, and do not clear usage/replay when the provider reports a skipped compact. See `docs/crewcoder-provider.md` and `docs/conversation-storage.md`.
Copy file name to clipboardExpand all lines: docs/current-state.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -147,7 +147,7 @@ Queued follow-up messages are also session-scoped. There is exactly ONE follow-u
147
147
148
148
Composer execution modes are `ask | plan | build | full`, where `full` is displayed as **Full Access**. A mode has an optional prompt preamble (`buildModePreamble`) **plus** a real per-provider permission policy; the preamble is advisory, not enforcement. Settings customizes the shared prompt text, while `Session.modePromptsEnabled` controls injection per Solo Chat from `ChatHeader`; it defaults true, copies on duplicate, and locks after startup context is sent because provider context cannot be revoked. Disabling it must never weaken provider-native permission policy, and skills/delegation context must still be delivered. Build approvals may grant **allow all (this turn ONLY)** only through main-issued `{bridgeId, turnId}` capability state: approve the current and later permission requests in that turn, clear on every turn/bridge termination boundary, and never use provider-native remembered choices that could outlive the turn. The wire value is `'full'` (renamed from `'yolo'`), the display token is `Mode` = `'Full'`, and the label is `MODE_LABEL` = `'Full Access'`. Because the wire value changed, persisted sessions and `defaultMode` can still hold `'yolo'`, so every read of a persisted mode goes through `normalizeModeLevel()` (`yolo` -> `full`, unknown -> `build`) — an unrecognized level returns `undefined` from `MODE_FROM_SETTINGS`, which crashes the composer mode picker and prepends a literal "undefined" to the prompt. Full Access means every tool is pre-approved: Claude uses `bypassPermissions`, Codex uses `workspace-write` + `never` approvals (deliberately NOT `danger-full-access` — worktree isolation stays), pi/hermes/CrewCoder auto-accept permission requests, and agents must never ask for tool permission. `toolPolicy: 'read-only'` always overrides the mode, so supervisor and completion bridges stay read-only even in Full Access, and real `AskUserQuestion` prompts still surface. See `docs/execution-modes.md`.
149
149
150
-
Reasoning effort choices are provider-native: Claude exposes `off/low/medium/high/xhigh/max`; Codex exposes `off/low/medium/high/xhigh/max/ultra`; other providers expose `off/low/medium/high/xhigh`. Claude `off` must explicitly disable SDK thinking, and its named levels map to SDK `effort`. Codex `xhigh/max/ultra` must pass through app-server unchanged rather than silently downgrade. Codex Ultra's automatic delegation is upstream provider behavior—CrewCode does not emulate it. See `docs/reasoning-effort.md`.
150
+
Reasoning effort choices are provider-native: Claude exposes `off/adaptive/low/medium/high/xhigh/max`; Codex exposes `off/low/medium/high/xhigh/max/ultra`; other providers expose `off/low/medium/high/xhigh`. Claude `off` must explicitly disable SDK thinking; `adaptive` omits both SDK `effort` and `thinking` to retain Claude Code's settings/model resolution, and its named levels map to SDK `effort`. Codex `xhigh/max/ultra` must pass through app-server unchanged rather than silently downgrade. Codex Ultra's automatic delegation is upstream provider behavior—CrewCode does not emulate it. See `docs/reasoning-effort.md`.
0 commit comments