@artooi/ag-ui-web-component 0.13.0 → 0.14.1

package/CHANGELOG.md

@@ -7,6 +7,74 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.14.1] — 2026-08-07
+
+### Security
+
+- **Five advisories closed**: `vite` → 7.3.6 (HIGH + MEDIUM), `postcss` → 8.5.26
+  (HIGH), `brace-expansion` → 2.1.4 (HIGH), `esbuild` → 0.28.1 (LOW). All four
+  are transitive, so they are pinned through `overrides` — there is no direct
+  dependency to bump.
+
+  ⚠ **`overrides` now live in `pnpm-workspace.yaml`**, not the `pnpm` field in
+  `package.json`; pnpm 11 ignores the latter and only warns. And an override
+  must be scoped to its major — an unbounded `brace-expansion: ">=2.1.2"`
+  resolves to 5.x, whose export shape `minimatch` cannot call, breaking `glob`
+  at runtime.
+
+- ⛔ **Four `dompurify` advisories are knowingly left open** (three LOW, one
+  MEDIUM), and the dependency is now pinned to **exactly `3.4.7`** rather than
+  `^3.4.7`.
+
+  3.4.8 through 3.4.13 **mis-sanitise under happy-dom**: `<script>` and `<img>`
+  pass straight through, which `tests/render_markdown.test.ts` catches. Verified
+  again against dompurify 3.4.13 with happy-dom 20.11.1 — moving the test DOM
+  forward does not fix it.
+
+  ⭐ **The pin was previously a caret range**, so the hold existed only in the
+  lockfile and nowhere in the manifest, undocumented — a `pnpm update` would
+  have silently disabled sanitisation. It is now exact, explained at the import
+  site in `src/ui/render_markdown.ts`, and recorded in `CLAUDE.md`.
+
+  The trade is deliberate: those advisories describe narrow bypasses, while
+  taking them costs *all* sanitisation under the only DOM the suite can run in.
+  `tests/render_markdown.test.ts` is the acceptance check for lifting it.
+
+## [0.14.0] — 2026-07-28
+
+### Added
+
+- **Run notices — a muted inline annotation for things the *run* did**, as
+  opposed to work the user asked for. Styleable via the `run-notice`,
+  `run-notice-icon` and `run-notice-text` parts; announced with `role="status"`
+  so a screen reader hears it politely rather than mid-sentence.
+  - **Compaction.** A standard AG-UI `ACTIVITY_SNAPSHOT` with
+    `activityType: "compaction"` (emitted by django-ag-ui 0.26+ when a
+    compaction capability trimmed the history) renders as "earlier turns
+    condensed", with the count. Activity events of any other type pass through
+    untouched, so another producer on that channel isn't mistaken for one.
+  - **Agent skills.** There is no dedicated event: loading a deferred capability
+    *is* an ordinary `load_capability` tool call, which is what reaches the
+    client. It now renders as `Using skill <id>` **instead of** the raw tool
+    card it would otherwise produce.
+
+    ⚠ Not to be confused with the existing `Skill` catalog — that is a *human*
+    affordance (a prompt the user launches from the chip row or `/`-palette).
+    An agent skill is chosen by the model mid-run. Only the latter emits a
+    notice.
+  - **Suppression covers all three paths a card can come from** — the live
+    stream, the client's tool-execution loop, and **restored history** — so a
+    reload shows the same transcript rather than resurrecting the raw
+    `load_capability` card.
+  - A `load_capability` call with **no usable id** falls back to a normal tool
+    card rather than being dropped: a malformed call is still real activity, and
+    hiding it would be worse than showing it plainly.
+- **`onActivity(activityType, content)`** on `AgUiClientHandlers`, forwarding
+  AG-UI activity events to the host element. `COMPACTION_ACTIVITY_TYPE` and
+  `LOAD_CAPABILITY_TOOL` are exported from `constants`.
+- **Two new overridable strings**: `historyCompacted` (token `{count}`) and
+  `usingSkill` (token `{name}`).
+
 ## [0.13.0] — 2026-07-27
 
 ### Added
@@ -578,7 +646,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Notes
 - First release — exercising the automated npm OIDC publish pipeline end-to-end.
 
-[Unreleased]: https://github.com/Artui/ag-ui-web-component/compare/v0.13.0...HEAD
+[Unreleased]: https://github.com/Artui/ag-ui-web-component/compare/v0.14.1...HEAD
+[0.14.1]: https://github.com/Artui/ag-ui-web-component/compare/v0.14.0...v0.14.1
+[0.14.0]: https://github.com/Artui/ag-ui-web-component/compare/v0.13.0...v0.14.0
 [0.13.0]: https://github.com/Artui/ag-ui-web-component/compare/v0.12.0...v0.13.0
 [0.12.0]: https://github.com/Artui/ag-ui-web-component/compare/v0.11.0...v0.12.0
 [0.11.0]: https://github.com/Artui/ag-ui-web-component/compare/v0.10.0...v0.11.0

package/README.md

@@ -48,6 +48,7 @@ No framework, no Django, no admin specifics live here. Downstream consumers (e.g
 - [New chat and collapse](#new-chat-and-collapse)
 - [Tool-call display modes](#tool-call-display-modes)
 - [Markdown rendering](#markdown-rendering)
+- [Run notices: compaction and agent skills](#run-notices-compaction-and-agent-skills)
 - [Skills: prompt chips and slash palette](#skills-prompt-chips-and-slash-palette)
 - [MPA durability: surviving full page reloads](#mpa-durability-surviving-full-page-reloads)
 - [Host seams: the SPA story](#host-seams-the-spa-story)
@@ -517,6 +518,37 @@ reveal). It honours `prefers-reduced-motion` (collapsing to instant).
 
 ---
 
+## Run notices: compaction and agent skills
+
+Some things a run does are neither text nor a tool the user asked for — the server condensed
+earlier turns to fit the context window, or the model pulled in an agent skill. Those render as
+**run notices**: a muted one-line annotation inline in the transcript, styleable via the
+`run-notice`, `run-notice-icon` and `run-notice-text` `part`s.
+
+> **Two different things are called "skills".**
+> The [prompt chips and slash palette](#skills-prompt-chips-and-slash-palette) below are a
+> **human** affordance — prompts *the user* launches. An **agent skill** is a folder of
+> instructions *the model* chooses to load mid-run. Only the second produces a run notice.
+
+Neither notice needs configuration here — both appear when the server is set up to produce them.
+
+**Compaction.** `django-ag-ui` emits a standard AG-UI `ACTIVITY_SNAPSHOT` with
+`activityType: "compaction"` when a compaction capability trimmed the history; the notice reports
+how many messages went. Server side, that means wrapping the capability in `CompactionObserver` —
+see [django-ag-ui's compaction guide](https://artui.github.io/django-ag-ui/compaction/). Activity
+events of any other type pass through untouched, so another producer on that channel is not
+mistaken for a compaction.
+
+**Agent skills.** There is no dedicated event for these: loading a deferred capability *is* an
+ordinary `load_capability` tool call, which is what reaches the client. The component recognises
+it, renders `Using skill <id>`, and suppresses the raw tool card that would otherwise appear
+beside it — on the live stream and on restored history alike, so a reload shows the same
+transcript. A `load_capability` call with no usable id falls back to a normal tool card rather
+than being dropped, since it is still real activity.
+
+Both strings are overridable like every other — `historyCompacted` (token `{count}`) and
+`usingSkill` (token `{name}`).
+
 ## Skills: prompt chips and slash palette
 
 Skills are pre-defined prompts the user can launch from a chip or the `/`-command palette. They are