neurosquad@0.6.0

Minor Changes

• #76 7383db9 Thanks @neurosquad-dev-bot! - The dashboard selects text and opens links itself. It holds the mouse, so the terminal couldn't do either: now a drag over an agent's text copies it (OSC 52, plus the system clipboard when local) and shows it inverted, and a click on an http(s) link opens it in the browser (copied instead over SSH). term-view gains linkAt, selectionText, bufferTop and a tile selection option.

Patch Changes

• Updated dependencies [7383db9]:
  • @neurosquad/term-view@0.2.0

neurosquad@0.5.3

Patch Changes

• #74 b70076e Thanks @neurosquad-dev-bot! - Prices GPT-6 Astra, GPT-6 Sol, GPT-6.1 Sol and GPT-6 Luna (OpenAI's list prices, with their rates above 272K prompt tokens). Codex's codex-auto-review and gpt-5.3-codex-spark have no published price and stay "no price". The price table is now dated 2026-10-13.
• Updated dependencies [b70076e]:
  • @neurosquad/core@0.4.2

@neurosquad/term-view@0.2.0

Minor Changes

• #76 7383db9 Thanks @neurosquad-dev-bot! - The dashboard selects text and opens links itself. It holds the mouse, so the terminal couldn't do either: now a drag over an agent's text copies it (OSC 52, plus the system clipboard when local) and shows it inverted, and a click on an http(s) link opens it in the browser (copied instead over SSH). term-view gains linkAt, selectionText, bufferTop and a tile selection option.

@neurosquad/core@0.4.2

Patch Changes

• #74 b70076e Thanks @neurosquad-dev-bot! - Prices GPT-6 Astra, GPT-6 Sol, GPT-6.1 Sol and GPT-6 Luna (OpenAI's list prices, with their rates above 272K prompt tokens). Codex's codex-auto-review and gpt-5.3-codex-spark have no published price and stay "no price". The price table is now dated 2026-10-13.

neurosquad@0.5.2

Patch Changes

• #72 ccb686f Thanks @neurosquad-dev-bot! - Prices Claude Sonnet 5.5 and Haiku 5.5 (Haiku 5.5 by prompt length: higher prices over 100,000 prompt tokens). The price table is now dated 2026-10-12.
• Updated dependencies [ccb686f]:
  • @neurosquad/core@0.4.1

neurosquad@0.5.1

Patch Changes

• #70 8d8373f Thanks @neurosquad-dev-bot! - Quiet while working: every way input reaches an agent now ends a quiet episode — typing, a paste, nsq send and a prompt from the phone, a queued or first prompt, an interrupt, and an answer's y/a/n keys (the dashboard, nsq answer, the phone), which before went in without restarting the quiet clock. They all go through one path in the daemon, so a new one cannot miss it. remote: PhoneHost.submit gets the caller, as answer and interrupt do.
• Updated dependencies [8d8373f]:
  • @neurosquad/remote@0.5.1

neurosquad@0.5.0

Minor Changes

• #68 3c14b5c Thanks @neurosquad-dev-bot! - Quiet while working: some prompts come from tools that never fire a hook — git asking for a password or a passphrase, an installer's [y/N], sudo, ssh's host-key question — or a command simply hangs. When a working agent's screen has not changed for quietAfter minutes (default 5, 0 = off, nsq config set quietAfter 10; per agent nsq set <agent> --quiet-after <min>), nsq flags it quiet: quiet 7m — check it in the tile, the sidebar and the header counter (1 needs you · 1 quiet), in nsq ls (and the quiet field of --json) and on the phone page, with the same desktop notification, terminal bell and ntfy push as "needs you" (<agent> has been quiet for 7 min plus the last screen line, secrets masked, at most 300 characters) — but no y/a/n keys. When the last lines look like a prompt (Password:, [y/N], (yes/no), a passphrase…) it says so: looks like it's asking: Password for 'https://github.com':. New output, typing or pasting into the agent, a sent prompt, an interrupt or a hook clears it; a turning spinner, a rotating verb or a counting timer does not count as output. Quiet episodes go into the approvals log ("kind": "quiet", start, end and what ended it). core: QuietWatch, screenChanged, quietPromptHint; remote: PhoneAgentSummary.quiet and the quiet attention kind; tui-theme: the quiet role.

Patch Changes

• Updated dependencies [3c14b5c]:
  • @neurosquad/core@0.4.0
  • @neurosquad/remote@0.5.0
  • @neurosquad/tui-theme@0.2.0

neurosquad@0.4.2

Patch Changes

• #65 146e508 Thanks @neurosquad-dev-bot! - The dashboard scrolls an agent's output back: the mouse wheel over a tile or a full-screen agent (3 lines a notch, a page with Shift) and Shift+PgUp / PgDn / Home / End. The view stays put while new output arrives and says how far back it is (↑ 120 lines · End or any key to return); End or any key for the agent returns to the live screen. Agents that asked for the mouse (OpenCode) or run full screen on the alternate screen still get the wheel and those keys themselves. The wheel over the sidebar keeps moving the selection. term-view: createScrollback pins the scrolled view with a marker; wheelTarget / wheelKeys route a wheel notch.

• #67 5082c6d Thanks @neurosquad-dev-bot! - nsq update and nsq doctor no longer list themselves as "1 nsq window is open" among the things holding the restart onto an installed update.

• Updated dependencies [146e508]:

  • @neurosquad/term-view@0.1.1

neurosquad@0.4.1

Patch Changes

• #63 3c06fb8 Thanks @neurosquad-dev-bot! - A newer nsq on an older running daemon (the hand-over waits while agents are busy) no longer shows a bare "unknown request". The daemon's hello now lists what it can do (capabilities) and its "unknown request" reply carries its version; clients check before they ask (older daemons by their version). The dashboard, its phone panel and every command say which daemon runs and what the request needs — "Press U to restart it now (agents resume their sessions), or it restarts by itself when agents are free" in the dashboard, nsq down && nsq up on the command line. The phone panel's Notifications tab explains it instead of failing on s, and offers U right there: the daemon restarts on this nsq and the panel opens again on the same tab. U in the dashboard also offers that restart whenever the daemon is older, not only while a hand-over waits.

@neurosquad/tui-theme@0.2.0

Minor Changes

• #68 3c14b5c Thanks @neurosquad-dev-bot! - Quiet while working: some prompts come from tools that never fire a hook — git asking for a password or a passphrase, an installer's [y/N], sudo, ssh's host-key question — or a command simply hangs. When a working agent's screen has not changed for quietAfter minutes (default 5, 0 = off, nsq config set quietAfter 10; per agent nsq set <agent> --quiet-after <min>), nsq flags it quiet: quiet 7m — check it in the tile, the sidebar and the header counter (1 needs you · 1 quiet), in nsq ls (and the quiet field of --json) and on the phone page, with the same desktop notification, terminal bell and ntfy push as "needs you" (<agent> has been quiet for 7 min plus the last screen line, secrets masked, at most 300 characters) — but no y/a/n keys. When the last lines look like a prompt (Password:, [y/N], (yes/no), a passphrase…) it says so: looks like it's asking: Password for 'https://github.com':. New output, typing or pasting into the agent, a sent prompt, an interrupt or a hook clears it; a turning spinner, a rotating verb or a counting timer does not count as output. Quiet episodes go into the approvals log ("kind": "quiet", start, end and what ended it). core: QuietWatch, screenChanged, quietPromptHint; remote: PhoneAgentSummary.quiet and the quiet attention kind; tui-theme: the quiet role.