Skip to content
cccskills
FreePublished on GitHub

context-menu-v3

The universal v3 right-click menu (EditableContextMenu / NonEditableContextMenu). Use when adding a right-click menu to a surface, fixing a fake menu with no Copy or Export, a missing Download as Markdown, or a bespoke per-surface menu. NOT for an assigned rollout shard (use context-menu-rollout).

The English summary is being prepared. Showing the original description.

See how to install

Files included (1)

  • SKILL.md10.3 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

context-menu-v3 — wire the universal menu onto a surface

<!-- rules:stamp -->

Rules from common-docs/systems/platform/ui-shell/FEATURE.md — edit there, never here (python3 common-docs/meta/scripts/one_home.py --write).

  1. Delete means archive, everywhere. A delete that does not archive is a defect.
  2. A menu regroup ships as a side-by-side demo page with a what-goes-where list, never as a direct change.
  3. Every surface mounts the canonical menu. Use NonEditableContextMenu or EditableContextMenu from features/context-menu-v3/. Never use a bespoke DropdownMenu or a raw contextmenu handler, and never keep both.
  4. A menu without a source is a defect. A menu with no contentSource, entity or surfaceName opens with every verb dark. Fix these before unwired surfaces.
  5. Lists built on EntityListPage inherit the row menu through ItemContextMenu. They need their own wiring only for surface-specific sections.
  6. A window or overlay mounts its own menu. It never delegates to the page under it.
  7. Reuse the domain's action registry. Never create a second one.
  8. Secret-bearing surfaces follow VaultContextMenu. Pass getApplicationScope, force selection: "", put names and field keys in the content (never values or notes), and pass no entity.
  9. Done means the menu opens clean. The console shows no INERT MENU or VALUE MAPPING GAP scream.
  10. Chat is a universal agent host. Never bind a page-agent roster to it.
  11. Anything with an id shows it through the one canonical id component (display plus one-click copy).
  12. Flagship and first-run screens get a bespoke elevated stage, not the 28px app-chrome controls.
  13. Before a link goes to Arman, every screen on its path passes the Steve Jobs bar (walked and fixed by an Opus reviewer).
  14. Admin tables filter, search and sort on the server, and people are picked by name or email, never by id.
  15. Wasted vertical space is the worst sin; every status shown is backed by real state.
  16. A new output style or menu layout is a NEW function used only by the new path; never thin a shared builder.
  17. A page fetches only what it shows, each thing once; every removed fetch ships its on-engagement load.
<!-- rules:stamp-end -->

The deep contract is features/context-menu-host/FEATURE.md — read it once. This skill is the rollout recipe; editing the v3 internals is that FEATURE.md's job, and a worker assigned a shard from pnpm check:context-menu follows the context-menu-rollout skill on top of this one. The proven reference is /demos/context-menu/canonical (every panel is a real v3 wiring; copy the one that matches your surface). The v2 menu (UnifiedAgentContextMenu) was deleted 2026-07-19 — every surface renders v3.

Pick the wrapper

  • EditableContextMenu — a textarea / editor. Gives Cut / Paste / Insert / Save / Delete on top of everything else — and auto-registers a WidgetHandle from the same callbacks, so agents launched from the menu can stream widget_text_* edits into the surface in place (nothing extra to wire; proof at /demos/context-menu/inline-edit).
  • NonEditableContextMenu — a viewer / rendered display / read-only text. No text mutation; Copy / AI / Export / Download / Convert still work because the menu self-resolves content from the DOM.

A surface with both modes (editor + preview) uses both — one per mode.

🚨 THE DENSITY LAW — labels only, macOS-terse (Arman, 2026-08-25)

A menu item is a short verb phrase: Edit rule…, See its keywords, Revert to pack…. No description / subtext, ever, with ONE exception: a disabled item carries a description naming why it's off and where it works ("Unchanged from the pack", "Works on the Keyword Workbench"). Rule of thumb: if macOS wouldn't put it in a menu, neither do we. Any other description requires Arman's explicit approval, requested AFTER the menu ships bare — and ~95% will be refused. An agent that ships prose under menu rows has failed the task.

🚨 CANONICAL SECTIONS — check the registry BEFORE writing any extraSections

features/context-menu-host/SECTIONS.md is the registry of shared per-identity section builders (keyword, class, level, page, CRM row, …) and THE ADOPTION PROTOCOL. The short form: if the right-clicked thing already has a registered builder, USE it — and grow it with the actions this surface makes obvious (every consumer gains them). If the thing appears on 2+ surfaces and has no builder, extract + register one. Only a truly page-local identity gets inline extraSections. Actions impossible on this surface stay visible-but-disabled with the reason as the tooltip — the menu is identical everywhere; only availability changes.

Wire a surface — the mechanical steps

  1. Import the wrapper statically — it's the light shell; children render synchronously, and only MenuContent lazy-loads on first open. Never wrap it in a per-consumer next/dynamic (that re-adds a loading fallback and a layout-collapse null state for zero benefit).
    import { EditableContextMenu } from "@ai-matrx/context-menu/react";
    
  2. Wrap the region in <EditableContextMenu …> (or NonEditableContextMenu).
  3. Pass the identity + value props: sourceFeature (required), surfaceName (registry surface → AI actions + bound agents + value mappings), getApplicationScope (live, preferred) / contextData (static), extraSections, and for editables getTextarea + onTextReplace / onTextInsertBefore|After; history props and scope / scopeId as the surface needs. Use placementMode for per-placement visibility. Types: @ai-matrx/context-menu/engine (EditableContextMenuProps / NonEditableContextMenuProps). sourceFeature names the actual product, not the nearest allowed label. If it is missing, register the product in aidream/aidream/services/conversation_context/source_attribution.py, regenerate the contract with its generator, and deploy the registry before the caller. Never cast or misattribute the caller to satisfy the allow-list.
  4. Build scope through buildApplicationScopeFromMenuContext (@ai-matrx/context-menu/engine) — it guarantees the 5 baselines from the live DOM.

ONE MENU PER PANE — delegate per row, never nest

A list, table, or grid gets one wrapper around the whole pane, not one per row. Nesting Radix triggers opens two menus and appears nowhere in this repo. Per-row context comes from resolveContextOnOpen(target): the shell calls it with the right-clicked element before opening, and the returned object is merged over contextData, so the same single menu can say Edit "China" on a row and show list-level rows on empty space. Worked reference: features/data-tables/pick-lists/components/ListDetailClient.tsx + its dom-anchors.ts (read the clicked row/group off data-* attributes).

⚠️ className on the wrapper styles the menu POPUP, not the trigger. Layout classes there silently break the popup instead of the pane — style the child element you wrap.

⚠️ An overlay/window surface must mount its own menu. Without one, a right-click inside the window is answered by the page underneath, handing the user THAT page's surface, values and agents — silently wrong, and it looks like it works.

Unlock the new capabilities (do this, don't skip it)

  • contentSource?: ContentSource (rich-document source: {type:"note",noteId}, {type:"chat-message",…}, {type:"raw"} default) → lights up Copy-as variants, Export, Download as Markdown, Convert and links Convert→Task to the right parent. A surface with a real entity should pass its source, not raw.
  • entity?: {type,id,title,resourceType?,isOwner?} → lights up Attach To (scope tagging) and Share. Omit on raw fragments.

Verify — non-negotiable

  • pnpm type-check clean.
  • Open the menu on the surface and watch the console. v3 SCREAMS in dev if the menu opens inert ("INERT MENU on …") or a surface dropped a declared value ("VALUE MAPPING GAP on …"). A clean open = the values are wired. A scream = fix the wiring (provide getApplicationScope/contextData.content, or make the wrapped content selectable for the DOM fallback).
  • For controlled editable fields, use the menu to replace text and confirm the owning state and derived output change together. A menu that changes only the DOM is not wired. Each field needs the callbacks for its own state.
  • The acceptance test for any content surface: right-click without selecting → Export → Download as Markdown saves the whole content as .md; highlight first → saves the selection.

Layout & density knobs (2026-08-22)

The desktop menu is model-driven: model/menu-model.ts (what exists) → model/layouts.ts (classic | tiered | command) → MenuContent.tsx (density comfortable | compact). Pass menuLayout / menuDensity on the wrapper only when a surface must deviate; the platform defaults are the CAPS constants in types.ts. A surface's extraSections need NO layout awareness — a long section folds into one submenu named by its label (+ optional icon) in tiered/command. Never add a layout by writing JSX in the renderer — add an arrangement in layouts.ts. 🚨 THE LOSSLESS LAW: a layout may never hide, rename, or drop a row Classic shows; grouping needs Arman's explicit approval (History is the only one today). Compare all four on /demos/context-menu/layouts.

Doctrine

  • Reuse, never fork. Every action (copy/export/convert/print/attach/share/AI) delegates to an existing system. A surface contributes its own items via extraSections (declarative anchors), never a bespoke menu.
  • One menu. No per-surface context-menu component. If a surface still has one, collapse it into a direct Editable/NonEditableContextMenu usage.

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

Admin debug system wiring for the floating AdminIndicator's live debug data and Copy Full Context. Use when adding debug visibility to a route or feature, wiring useDebugContext or the debug panel, capturing console errors, or enabling the copy-context workflow for a page.

The English summary is being prepared. Showing the original description.

armanisadeghi/ai-matrx3Updated Oct 20, 2026

Compact two-icon Copy / Copy-for-AI controls (components/agent-copy). Use when adding copy buttons to a row, card, list, or record; merging duplicate Copy/JSON/AI controls; continuing the copy rollout; or writing a Copy-for-AI payload. NOT for markdown content actions (use rich-document-actions).

The English summary is being prepared. Showing the original description.

armanisadeghi/ai-matrx3Updated Oct 20, 2026

Disclosing a surface's existing fixed AI jobs in the shell's top Agents menu. Use when a page, panel, overlay, or window already runs a mandate behind a button, assist, automatic action, or mode; when the agent-disclosure guard names a file; or during a surface check.

The English summary is being prepared. Showing the original description.

armanisadeghi/ai-matrx3Updated Oct 20, 2026

The watch-fix-rerun method for making an existing platform agent or automated process efficient and correct: baseline its ledger, run one unit yourself, fix the class behind every wasted call, rerun, record. Use when asked to improve, optimize, watch, or 'make efficient' an agent, a sandbox session, a sync job, or any recurring automated process.

The English summary is being prepared. Showing the original description.

armanisadeghi/ai-matrx3Updated Oct 20, 2026

Redux state for agent execution and firing agent shortcuts. Use when editing features/agents/redux/, building agent UI, creating a conversation, touching assembleRequest or NDJSON stream state, wiring agent-state selectors, adding a per-conversation capability, or triggering a shortcut from a button, menu, or code (useShortcutTrigger, launchAgentExecution).

The English summary is being prepared. Showing the original description.

armanisadeghi/ai-matrx3Updated Oct 20, 2026

The provision for a call site: the exhaustive menu of values that place in the code can realistically produce. Use before creating or fixing a mandate or agent that reasons about more than its own input, when deciding whether an agent can answer at all with what it is sent, or when its output 'looks right' but nobody checked it could be. NOT for building the agent (use create-agent).

The English summary is being prepared. Showing the original description.

armanisadeghi/ai-matrx3Updated Oct 20, 2026

See all skills by armanisadeghi

Report a problem with this skill