GeiserX · Nov 3, 2026 · Nov 2, 2026 · Nov 2, 2026
diff --git a/.claude-plugin/plugin.json b/.claude-plugin/plugin.json
@@ -1,6 +1,6 @@
 {
   "name": "akou",
-  "version": "1.7.4",
+  "version": "1.7.5",
   "description": "Record your calls locally with akou and question them while they run: the akou skills and the akou_* MCP tools.",
   "author": { "name": "GeiserX" },
   "homepage": "https://github.com/GeiserX/akou",

diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -2,6 +2,17 @@
 
 All notable changes to akou. Versions follow [semantic versioning](https://semver.org); the 0.x releases were prereleases, and from 1.0.0 on a release is a full release unless its version has a prerelease part, such as 1.1.0-rc.1.
 
+## 1.7.5 — the menu bar item on a Mac is no wider than its neighbours
+
+On a Mac, akou's item in the menu bar was 48 pt wide, with blank space beside the mark, while the items around it are 24 to 40 pt. It is now 32 pt, and the mark stays sharp on a Retina display. Windows and Linux are unchanged. This release also records a new reading of the Anthropic and OpenAI terms, in which nothing changed ([docs/providers.md](docs/providers.md#the-reading-of-2026-11-02)). The 8-hour soak did not run for this release: it exercises the call-recording helper, and nothing in this release touches it.
+
+### Tray and menu bar
+- **The macOS menu bar item is 32 pt wide, not 48 pt** (#508). The tray library sizes the item for the icon as the file loads, then shrinks the icon to 16 pt, and the item keeps the first width. Our 32 px icon file gave no resolution, so it loaded at 32 pt. Both macOS icons now say 144 dpi, so they load at 16 pt from the start.
+
+### Behind the scenes
+- The nightly's remote-dictation stage now starts each test dictation where a speech clip starts, not inside a word; remote dictation did not regress with sherpa-onnx 2.0.1 (#507).
+- The soak now says when its worst tick was late and how much CPU akou used then (#506).
+
 ## 1.7.4 — a microphone that is slow to wake no longer swallows a dictation
 
 On a Mac, a USB microphone waking from idle can take more than 3 seconds to open. akou gave up on it after 3 seconds and threw the stream away just before it arrived. The dictation key then seemed to do nothing, and nothing said why: the reason went only to `app.log`. Now akou keeps waiting for as long as the key is held or the dictation is latched on, and dictates from the moment the microphone opens. When it never opens, the dictation bar says "The microphone didn't open". A microphone that opens straight away works as before. Windows and Linux had the same 3-second limit in dictation and get the same fix. Recording a call is unchanged: it already waited longer and said when a device failed ([TRAPS](docs/TRAPS.md) T4.39). The 8-hour soak did not run for this release: it runs the call-recording helper from a file and opens no device, and this release changes only how dictation opens a device.

diff --git a/docs/api/openapi.json b/docs/api/openapi.json
@@ -2,7 +2,7 @@
   "openapi": "3.1.0",
   "info": {
     "title": "akou",
-    "version": "1.7.4",
+    "version": "1.7.5",
     "description": "The akou API over HTTP. Every operation is marked with the modes that serve it (`x-akou-modes`: the desktop app, server mode, or both) and the key it needs (`x-akou-access`: `open`, a scope, or the list of scopes that reach it; an `admin` key reaches every operation). Generated from the route table; do not edit by hand.",
     "license": {
       "name": "GPL-3.0-or-later",

diff --git a/docs/providers.md b/docs/providers.md
@@ -166,6 +166,11 @@ Before each release we read the current terms and the tools' own documentation,
 | 2026-10-29 | Allowed as akou runs it, on our reading; Anthropic has not confirmed it. Nothing changed since the second reading of 2026-10-28, made for 1.4.1: the Claude Code legal page, read directly, is identical in its text to the web archive copies of 27 and 29 October 2026 and still says Anthropic does not permit third-party developers "to offer Claude.ai login into their own applications, or to route requests through Free, Pro, or Max plan credentials on behalf of their users", still allows an end user signing in "to the unmodified Claude Code binary with their own Claude subscription", and still says Pro and Max limits "assume ordinary, individual usage"; the Usage Policy, read directly, is still the one effective 12 November 2026, with the clause on services "that route requests through consumer subscriptions"; the Consumer Terms, read directly, are still the version effective 8 October 2025, with the same automated-access and credential-sharing clauses. The new `command` provider does not change this: it runs a program the user names in their own config file, under whatever login that program uses, and akou picks no account and handles no credential | Allowed as akou runs it, on our reading; OpenAI has not confirmed it. Nothing changed since the second reading of 2026-10-28: the Usage Policies, read in a browser, are still the version effective 20 November 2026, with no changelog entry after 2026-10-21, with the same model-access intermediary paragraph and its exemption for "an application integrates OpenAI models to power its features or functionality"; the Terms of Use effective 1 January 2026 and the Europe Terms of Use updated 16 January 2026, read in a browser, keep the clauses on extracting Output "automatically or programmatically" and on sharing account credentials; the Codex non-interactive page, read directly, still says `codex exec` reuses saved CLI authentication by default, still offers `--json` for scripts and still says `--full-auto` is no longer supported. Same open question, and the `command` provider does not change it either | None. The harness stays the default, the `command` provider is opt-in from the config file only, and the rolling memo stays off for both. Both questions stay open |
 | 2026-10-30 | Allowed as akou runs it, on our reading; Anthropic has not confirmed it. One wording change since 2026-10-29, not a new rule: the Claude Code legal page, read directly, now names an Ultra plan beside Free, Pro and Max. It says Anthropic does not permit third-party developers "to offer Claude.ai login into their own applications, or to route requests through Free, Pro, Max, or Ultra plan credentials on behalf of their users", still says it does not "prevent an end user from signing in to the unmodified Claude Code binary with their own Claude subscription", and says "Advertised usage limits for Pro, Max, and Ultra plans assume ordinary, individual usage". Against the web archive copy of 29 October 2026 nothing else in its text differs. The Usage Policy, read directly, is still the one effective 12 November 2026, with the clause on services "that route requests through consumer subscriptions"; the Consumer Terms, read directly, are still the version effective 8 October 2025, with the same automated-access and credential-sharing clauses | Allowed as akou runs it, on our reading; OpenAI has not confirmed it. Nothing changed since 2026-10-29: the Usage Policies, read in a browser, are still the version effective 20 November 2026, with no changelog entry after 2026-10-21, with the same model-access intermediary paragraph and its exemption for "an application integrates OpenAI models to power its features or functionality"; the Terms of Use effective 1 January 2026 and the Europe Terms of Use updated 16 January 2026, read in a browser, keep the clauses on extracting Output "automatically or programmatically" and on sharing account credentials; the Codex non-interactive page, read directly, still says `codex exec` reuses saved CLI authentication by default, still offers `--json` for scripts and still says `--full-auto` is no longer supported. Same open question | None. The harness stays the default, the `command` provider is opt-in from the config file only, and the rolling memo stays off for both. Both questions stay open |
 | 2026-11-01 | Allowed as akou runs it, on our reading; Anthropic has not confirmed it. Nothing changed since 2026-10-30: the Claude Code legal page, read directly, is identical in its text to the web archive copy of 31 October 2026 and still says Anthropic does not permit third-party developers "to offer Claude.ai login into their own applications, or to route requests through Free, Pro, Max, or Ultra plan credentials on behalf of their users", still says it does not "prevent an end user from signing in to the unmodified Claude Code binary with their own Claude subscription", and still says "Advertised usage limits for Pro, Max, and Ultra plans assume ordinary, individual usage". The Usage Policy, read directly, is still the one effective 12 November 2026 and identical in its text to the web archive copy of 28 October 2026, with the clause on services "that route requests through consumer subscriptions"; the Consumer Terms, read directly, are still the version effective 8 October 2025, with the same automated-access and credential-sharing clauses | Allowed as akou runs it, on our reading; OpenAI has not confirmed it. Nothing changed since 2026-10-30: the Usage Policies, read in a browser, are still the version effective 20 November 2026, with no changelog entry after 2026-10-21, with the same model-access intermediary paragraph and its exemption for "an application integrates OpenAI models to power its features or functionality"; the Terms of Use effective 1 January 2026 and the Europe Terms of Use updated 16 January 2026, read in a browser, keep the clauses on extracting Output "automatically or programmatically" and on sharing account credentials; the Codex non-interactive page, read directly, still says `codex exec` reuses saved CLI authentication by default, still offers `--json` for scripts and still says `--full-auto` is no longer supported. Same open question | None. The harness stays the default, the `command` provider is opt-in from the config file only, and the rolling memo stays off for both. Both questions stay open |
+| 2026-11-02 | Allowed as akou runs it, on our reading; Anthropic has not confirmed it. Nothing changed since 2026-11-01: the Claude Code legal page, read directly after v1.7.4 was tagged, is identical in its text to the web archive copies of 31 October and 1 November 2026 and still says Anthropic does not permit third-party developers "to offer Claude.ai login into their own applications, or to route requests through Free, Pro, Max, or Ultra plan credentials on behalf of their users", still says it does not "prevent an end user from signing in to the unmodified Claude Code binary with their own Claude subscription", and still says "Advertised usage limits for Pro, Max, and Ultra plans assume ordinary, individual usage". The Usage Policy, read directly, is still the one effective 12 November 2026 and identical in its text to the web archive copy of 2 November 2026, with the clause on services "that route requests through consumer subscriptions"; the Consumer Terms are still the version effective 8 October 2025, the EEA one read directly and the one outside the EEA through a web archive copy of 30 October 2026, with the same automated-access and credential-sharing clauses | Allowed as akou runs it, on our reading; OpenAI has not confirmed it. Nothing changed since 2026-11-01: the Usage Policies, read in a browser, are still the version effective 20 November 2026, with no changelog entry after 2026-10-21, with the same model-access intermediary paragraph and its exemption for "an application integrates OpenAI models to power its features or functionality"; the Terms of Use effective 1 January 2026 and the Europe Terms of Use updated 16 January 2026, read in a browser, keep the clauses on extracting Output "automatically or programmatically" and on sharing account credentials; the Codex non-interactive page, read directly, differs from the web archive copy of 21 October 2026 only in its navigation links, and still says `codex exec` reuses saved CLI authentication by default, still offers `--json` for scripts and still says `--full-auto` is no longer supported. Same open question | None. The harness stays the default, the `command` provider is opt-in from the config file only, and the rolling memo stays off for both. Both questions stay open |
+
+### The reading of 2026-11-02
+
+We read the pages again for the 1.7.5 release, after v1.7.4 was tagged the same day. Nothing changed since the reading of 2026-11-01. The Claude Code legal page, read directly, matches the web archive copies of 31 October and 1 November 2026 sentence for sentence. The Usage Policy effective 12 November 2026, read directly, matches the web archive copy of 2 November 2026 sentence for sentence. The Consumer Terms effective 8 October 2025 keep the clauses quoted below, the EEA version read directly and the version outside the EEA through a web archive copy of 30 October 2026. openai.com still answers 403 to a plain fetch, so the Usage Policies (effective 20 November 2026, newest changelog entry 2026-10-21), the Terms of Use (effective 1 January 2026) and the Europe Terms of Use (updated 16 January 2026) were read in a browser, and each keeps the clauses quoted in the readings below. Codex's non-interactive page, read directly, differs from the web archive copy of 21 October 2026 only in its navigation links. The UK version of Anthropic's Consumer Terms is still not covered.
 
 ### The reading of 2026-11-01
 

diff --git a/examples/compose/telegram-archive/.env.example b/examples/compose/telegram-archive/.env.example
@@ -2,7 +2,7 @@
 # docker-compose.yml.
 
 # The drumsergio/akou version to run. There is no latest tag: name one.
-AKOU_VERSION=1.7.4
+AKOU_VERSION=1.7.5
 
 # Printed once by `akou keys create --name archive --scope jobs --callback-host telegram-viewer`:
 # the ak_ key for the backup, and the whsec_ secret the viewer checks each callback with.

diff --git a/native/akou-capture/Cargo.lock b/native/akou-capture/Cargo.lock
diff --git a/native/akou-capture/Cargo.toml b/native/akou-capture/Cargo.toml
@@ -1,6 +1,6 @@
 [package]
 name = "akou-capture"
-version = "1.7.4"
+version = "1.7.5"
 edition = "2024"
 rust-version = "1.88"
 license = "GPL-3.0-or-later"

diff --git a/native/akou-diarize/Cargo.lock b/native/akou-diarize/Cargo.lock
diff --git a/native/akou-diarize/Cargo.toml b/native/akou-diarize/Cargo.toml
@@ -1,6 +1,6 @@
 [package]
 name = "akou-diarize"
-version = "1.7.4"
+version = "1.7.5"
 edition = "2024"
 rust-version = "1.88"
 license = "GPL-3.0-or-later"

diff --git a/package.json b/package.json
@@ -1,6 +1,6 @@
 {
   "name": "akou",
-  "version": "1.7.4",
+  "version": "1.7.5",
   "private": true,
   "type": "module",
   "license": "GPL-3.0-or-later",

diff --git a/skills/akou-vocab/SKILL.md b/skills/akou-vocab/SKILL.md
@@ -2,7 +2,7 @@
 name: akou-vocab
 description: Teach akou the names and terms of the user's world, so recordings spell them right. Use before a call when the user shares an invite, when the user asks to prepare akou's vocabulary, or after a call to turn the words it got wrong into entries. Every word you find ends as a proposal the user approves; you never confirm one yourself.
 metadata:
-  version: "1.7.4"
+  version: "1.7.5"
 ---
 
 # akou-vocab

diff --git a/skills/akou/SKILL.md b/skills/akou/SKILL.md
@@ -2,7 +2,7 @@
 name: akou
 description: Record a call or meeting on this computer with akou and answer questions about it while it runs, or transcribe an audio file. Use when the user says record this call or meeting, starts a call, names a speaker, spells a word, asks what was said, decided or is being discussed, or hands over a recording or voice note to transcribe.
 metadata:
-  version: "1.7.4"
+  version: "1.7.5"
 ---
 
 # akou

diff --git a/src/main/app-info.ts b/src/main/app-info.ts
@@ -4,7 +4,7 @@
  */
 
 /** The one version, stamped from `package.json` at release (DESIGN section 9). */
-export const APP_VERSION = "1.7.4";
+export const APP_VERSION = "1.7.5";
 /** `runtime.json` in the config folder: pid, port and version of the running app, mode 0600. */
 export const RUNTIME_FILE = "runtime.json";
 /** `akou.lock` in the config folder: the pid of the one app, taken early in its start. */