v0.2.5 (early release)

Install or update:

• macOS / Linux: curl -fsSL https://enowx.ai/install.sh | sh
• Windows (PowerShell): irm https://enowx.ai/install.ps1 | iex

Each archive has a .sha256 checksum beside it.

Fixes

• Agents no longer sit forever on a command that never ends, on Windows
  especially. When a shell command timed out or was stopped, only Unix
  killed the whole process tree; on Windows the shell's children (a dev
  server, a watcher, a prompt waiting for input) kept running and enowx
  waited on them for good, so the agent looked stuck and Stop did nothing.
  The tree is now ended with taskkill /T /F, and the wait after a kill is
  bounded everywhere.
• A model stream that goes silent is retried after 3 minutes without a
  byte, instead of holding the turn until the request's 10-minute limit.
• A formatter that never exits no longer holds the edit that ran it; it
  is stopped after 60 seconds.
• A sub-agent that has made no progress for 5 minutes says so in the
  sidebar (quiet 7m), with a hint to right-click it to stop it.
• Stopping a sub-agent that is no longer running (stopped already, or
  left from an earlier run) marks it stopped instead of leaving it shown as
  at work.
• A session with a damaged line still opens and resumes: the line is
  skipped instead of refusing the whole session, and opening a sub-agent's
  transcript that cannot be read says why. On Windows, saving retries
  briefly when another program holds the file for a moment.

v0.2.4 (early release)

Install or update:

• macOS / Linux: curl -fsSL https://enowx.ai/install.sh | sh
• Windows (PowerShell): irm https://enowx.ai/install.ps1 | iex

Each archive has a .sha256 checksum beside it.

Fixes

• Quitting no longer freezes while a specialist is at work under an ACP
  lead. An ACP agent's delegation ran the specialist inside the lead's tool
  call, and quitting waited for that turn to end. Stopping the lead's turn
  now answers the call at once, quitting stops the specialists still at work,
  and the exit never waits more than a few seconds for a turn.
• The sidebar names the engine a specialist runs on. With /acp in use,
  every specialist runs on that agent, but its row showed enowx's configured
  model; it now reads Claude Code (ACP) and its model.
• enowx, not enx, everywhere you read it: the quit dialog, /update and
  its status lines, enowx update, and the messages that tell you which
  command to run. enx still works as a second name for the command.
• enowx dev runs the rebuilt enowx instead of a binary that no longer
  exists.

v0.2.3 (early release)

Install or update:

• macOS / Linux: curl -fsSL https://enowx.ai/install.sh | sh
• Windows (PowerShell): irm https://enowx.ai/install.ps1 | iex

Each archive has a .sha256 checksum beside it.

Agents

• The lead decides instead of asking. The orchestrator brainstormed almost
  every build request and asked which checks to run, so even a small, clear
  project started with questions. It now goes straight to work, fills open
  details from the request, the code and convention, and states its
  assumptions in one line. A brainstorm runs only when you ask for one, or
  when a large new product's purpose cannot be told from the request. It asks
  only what is yours to decide (taste nothing hints at, money, anything
  irreversible, credentials), once, with a recommendation. The check wave
  (review, tests, security where it matters) runs without a question, plan
  documents are written only when you ask for them, and inside a brainstorm
  the stack is decided, not asked. The maestro follows the same rules.

ACP agents

• Run the session on Claude Code, Codex, Gemini CLI, Kiro CLI or any ACP agent.
  /acp claude puts the lead and every specialist on it, after installing its
  adapter into ~/.enx/acp if needed (never globally) and checking the
  sign-in; names complete as you type. /model and /effort then list what
  that agent offers, read from it, chosen with the arrows and kept per agent;
  specialists use the same model. /acp off (or /native) goes back to
  enowx's own model, and /acp alone lists the agents with their status.
  The status bar always says which is in use, with its model, effort and
  permissions. Settings > ACP agents has a card per agent (status, default
  model and effort from its list, permissions, run the session on it) and a
  row to add a custom ACP agent. Permissions are ask me by default, enowx's
  rules, or bypass (only when chosen, with a warning). The agent gets the
  role's prompt, and enowx's skills, delegation to its specialists and
  questions to you over a local MCP server (127.0.0.1, a random bearer token
  per session); its messages, reasoning, tool calls, diffs and plan show as
  enowx's own, Esc cancels its turn, and it stays warm between turns. e in
  the roster puts one agent on another engine, for mixing. enowx acp status
  and enowx acp install <engine> work from a shell.
• Kiro CLI is built in. /acp kiro runs the session on kiro-cli acp
  (1.25 or newer, found on the PATH, never installed by enowx); its card
  shows the version and the account type signed in, without the email, and
  /model lists Kiro's models. It runs as enowx's own Kiro agent
  (~/.kiro/agents/enowx.json, never in the project, never over a file the
  user wrote), so it loads only enowx's MCP server instead of the ones in the
  user's Kiro config and their OAuth prompts. An agent's error now carries its reason (the
  model not available on the plan, say) instead of "Internal error".

Composer

• Earlier prompts with Up and Down, and Ctrl+W to erase a word. Up on
  the composer's first line brings back the prompts sent in this session,
  newest first, and Down on its last line steps forward again, back to what
  you were typing. Ctrl+W erases the word before the caret, as in a shell; it
  is also what Windows Terminal sends for Ctrl+Backspace, which other
  terminals report either as itself (already handled) or as a plain
  Backspace. Reported in #37.

Install and update

• enx and enowx are both the command. The installers put enx beside
  enowx (a link on macOS and Linux, a copy on Windows), and updating keeps
  both names on the new version whichever one ran. An older enowx or enx
  elsewhere on the PATH, such as one cargo install left in
  ~/.cargo/bin, is pointed at the new install, so the old version no longer
  starts; a program named enx that is not enowx is left alone.
• enowx update works again. It still asked for archives named
  enx-<target>, which v0.2.2 no longer published, so it failed. It now asks
  for enowx-<target>, and releases also publish enx-<target> with both
  names inside, so enx update from v0.2.1 and enowx update from v0.2.2
  reach the new version.
• Uninstall. ENOWX_UNINSTALL=1 with either install script removes both
  names. ~/.enx (settings, keys, sessions) is never touched.

Decision model

• A decision model for small typed judgements, off by default. The
  Decision model section of Settings (or /decision) picks a provider behind
  one interface:
  Clef on Cloudflare Workers AI, TypeSafe's Jev, any endpoint speaking the
  same API, or a model already configured in enowx asked for JSON. Five uses,
  each on its own switch and threshold: whether a message asks for a
  brainstorm, which specialist it is for, whether a question is the user's
  to answer, whether a long tool result is carried whole, and whether a shell
  command needs the user's word first (an unsure answer always asks). Every
  judgement has a hard timeout (300 ms by default) and falls back to the
  usual path when late or failing; shadow mode records decisions without
  acting on them. Decisions are logged to ~/.enx/decisions.jsonl, listed
  in the Log tab, and printed by enowx decisions.

Removed

• TypeSafe integration. The TypeSafe (System One) judge that trimmed tool
  results and kept live turns through compaction is gone, with its Settings
  section, the /typesafe command, the sidebar trim count and the Log tab's
  context filter. Tool results are always kept whole and compaction is the
  plain fold by age, which is what ran without a TypeSafe key. enowx no
  longer contacts TypeSafe. A [typesafe] table left in config.toml
  still loads, is ignored, and is kept as written when the file is saved;
  TYPESAFE_API_KEY is read only as the key of the Jev decision model, above.

v0.2.2 (early release)

Install or update:

• macOS / Linux: curl -fsSL https://enowx.ai/install.sh | sh
• Windows (PowerShell): irm https://enowx.ai/install.ps1 | iex

Each archive has a .sha256 checksum beside it.

Renamed to enowx

• The command is now enowx, not enx. The project name and the binary
  are enowx everywhere; enx auth login becomes enowx auth login, and so
  on. The install scripts, release archives (enowx-<target>), docs and the
  website use enowx. Your data in ~/.enx/ is left where it is, so settings,
  keys and sessions carry over; only the command you type changed. Old
  ENX_* environment knobs still work alongside the new ENOWX_* ones.

Models and agents

• A model manager that tests before it assigns. A new model-manager
  skill (carried by the orchestrator and maestro) helps set a model per agent:
  enowx models list shows the connected providers and their models, and
  enowx models test provider/model makes one tiny call and reports whether
  it answered, how fast, and the tokens. The skill lists what you own,
  brainstorms a mapping, tests every candidate, and sets only the ones that
  answered.
• The footer and /model follow the active agent's model. Setting an
  agent to another model shows at once; picking a model while an agent has its
  own changes that agent's model, not the conversation's.
• Each delegated sub-agent shows the model it runs on, in the sidebar list
  and the log.
• Stopping a sub-agent asks first (a right-click opens a confirm that
  defaults to keeping it).

Agents

• Brainstorm no longer asks how the work is built. It asks only what is
  yours to decide (purpose, scope, look, content) and reads the one relevant
  skill before offering options, so it never invents tool choices or asks you
  to do a step a tool can do.
• A motion-video skill: render a video file (an explainer, an animated
  logo) from code with synthesised sound, instead of reaching for video
  software or asking you to record.

v0.2.1 (early release)

Install or update:

• macOS / Linux: curl -fsSL https://enowx.ai/install.sh | sh
• Windows (PowerShell): irm https://enowx.ai/install.ps1 | iex

Each archive has a .sha256 checksum beside it.

Interface work

• Everything that can be set is in Settings. New sections: General
  (browser preview, language server checks, background delegation,
  switching agent by itself, compaction and its threshold, model calls per
  turn, the shell timeout, the model for each tier), Display (sidebar,
  currency and rate) and TypeSafe. Each field is saved through the same
  checks as enx config set, and a bad value is refused with why.

• Structure before styling: ui-structure and ui-anatomy. The
  frontend, canvas and mobile agents now write a blueprint before any
  code: the screen's job and its one primary action, the regions its kind
  of page needs in order (homepage, dashboard, list, detail, settings,
  form, sign-in, pricing, docs, article, checkout) with why each is there,
  the skeleton in numbers (container, gutters, grid, spacing, what each
  breakpoint changes), and the components with their states. They build
  in a fixed order (tokens, layout primitives, components, sections, the
  page) and every control meets its contract in ui-anatomy: parts,
  variants, sizes, every state, keyboard.

• Reviewers judge structure before taste. Content touching the
  window's edge, sections on different left edges, mixed surfaces, cut-off
  phone layouts, selected states shown by a shade, invalid submits and
  missing loading, empty or error states rule out Good; a page not seen
  in a browser is never called good to look at.

• preview reports more: content touching the window's edge (no
  gutter), content cut off past it, and a theme toggle with text on it.

Agents and delegation

• The orchestrator and maestro can stop a delegation at work with
  stop_delegation and a reason; its report says who stopped it and why,
  and the list of delegations still running now gives their sessions.
• Each delegated sub-agent shows the model it runs on in the sidebar
  list and the log, so a glance shows what is answering each one.
• Stopping a sub-agent asks first. A right-click in the sidebar used to
  stop it at once; it now opens a confirm that defaults to keeping it, so a
  stray click stops nothing.

Models

• A model manager that tests before it assigns. A new model-manager
  skill, carried by the orchestrator and maestro, helps set a model per
  agent: enx models list shows the connected providers and their models,
  the model in use and the per-tier and per-agent ones; enx models test provider/model makes one tiny call and says whether it answered, how
  fast, and the tokens, failing with the provider's own error. The skill
  lists what you own, brainstorms a mapping, tests every candidate, then
  sets only the ones that answered.
• The footer and /model follow the active agent's model. Setting an
  agent to another model used to leave the footer on the old one until the
  next /model, and picking a model while an agent had its own changed the
  conversation's model, not the agent's. Both now track the agent that is
  answering.
• antares --version and --help work as commands (the CLI took a
  leading flag as an option of the default serve).

v0.2.0 (early release)

Install or update:

• macOS / Linux: curl -fsSL https://enowx.ai/install.sh | sh
• Windows (PowerShell): irm https://enowx.ai/install.ps1 | iex

Each archive has a .sha256 checksum beside it.

Still an early release: expect rough edges, and please report them in
Issues. The highlights: settings move to a Settings tab beside the chat;
agents can work together (messages, a shared board, cross-review) when you
turn it on; built-in code search over your project (rag); built-in MCP
servers for Coolify, Dokploy and your VPSes; a stopped or failed sub-agent
can be read back by the agent that sent it; every shortcut now works on
Linux, macOS and Windows; and enx updates itself (enx update, /update).

Agents and delegation

• Sub-agents count, and are logged. What a delegated agent uses comes
  up to the interface as it works: the Log tab lists every agent's model
  calls, retries, trims and errors, named; the SESSION card's new used
  row and the cost count the whole session, sub-agents included (a
  reviewer's tokens too); and with a sub-agent's transcript open, the
  context shown is that agent's own.

• A finished sub-agent leaves the sidebar once its transcript is
  cleared, so no row leads nowhere; one at work, failed or stopped stays.

• One reviewer across cross-review rounds. It waits between rounds in
  its own session (so the delegate can answer it) and carries on there,
  keeping its context and the prompt cache.

• Stop a sub-agent with a right-click on it in the sidebar. Its caller
  gets a report that it was stopped by the user, with the files it had
  changed and what it said last, and is woken by it like any report; its
  transcript is kept.

• The orchestrator and maestro can read what a delegation did
  (delegation_log): what it was asked, said and called, so after one is
  stopped or fails the next brief starts where that work stopped.

• Agents can work together (Settings > Team, off by default). Agents at
  work at the same time message each other (message_agent), share a board
  for the run (team_board), and a delegate's work that changed files is
  cross-reviewed: the reviewer's corrections go back to the delegate and the
  work is checked again, up to the rounds set (1 to 5). Each part can be
  turned off on its own.

• New maestro agent: an all-rounder in the LEAD group. It has the
  orchestrator's reach (delegate and hand off to anyone) but carries every
  tool, so it can change files, run commands and check interfaces itself
  rather than only handing work out. For a job that mixes doing and
  delegating. The orchestrator stays the default; pick it with
  /agent maestro.

• A delegated agent always comes back with a report. The four fields
  (DONE, CHANGED, VERIFIED, NEXT) are now read however the model dresses them
  (markdown, changed case); every ending without a report is nudged; after two
  nudges the model is asked once more with no tools, so all it can do is write
  the report; and if it still will not, the caller gets the four fields built
  from what the branch actually did.

• A delegation at work is never run twice. Resuming a branch that is still
  running is refused, and the orchestrator is told up front what it left
  running, so a "continue" from the user can no longer start the same work a
  second time.

• A finished delegation is cleaned up. Once a delegation reports
  successfully its report stays in the conversation and its transcript (and any
  delegation under it) is deleted instead of filling the disk. One that failed
  or stopped short is kept, to be resumed.

• Reviewers are firm, tidy and honest. review and perf open with one
  verdict on a fixed scale (Good, Good with fixes, Needs work, Bad) that
  follows from the findings, call bad work bad and good work good with the
  reason, and treat inconsistency as a finding rather than a nit to drop.

• Give an agent its own model from the roster. In /agent, m opens the
  model list for the selected agent and d puts it back on the default.
  Every agent shows the model it runs on under its description.

Interface

• enx updates itself. enx update installs the latest release in place
  of the running binary after checking its published checksum (--check
  only looks, --force reinstalls). The interface looks for a new release
  in the background at start and notes it in the status bar until
  /update installs it. Settings > Updates turns the check off, or has it
  install by itself, used from the next start.

• Line edits show as diffs. An edit by line anchors used to show the
  tool's raw anchored text; it now returns a small diff the interface draws
  like any edit.

• Chat and Settings tabs at the top right. Settings takes the whole main
  column in place of the chat: its sections (Models, Providers, Agents, Team,
  MCP, RAG, Skills, Sessions, Theme) are listed on the left and the chosen one is
  beside them. Ctrl+P switches between Chat and Settings. Settings opens
  on the section list: Up/Down pick a section, Enter goes in, and Esc
  steps back one level (form to list, section to section list, list to
  chat); a click on a section picks it, a click inside goes in. The searchable command list
  moved to /commands (typing / still lists commands inline).

• Queue messages typed while a turn runs. Enter during a turn puts the
  message in a queue above the composer, with a [send now] button; queued
  messages go one at a time as each turn ends. Ctrl+Enter or Ctrl+S sends
  now, Shift+Enter or Alt+Enter is a newline at any time, and Up in an
  empty composer edits the last queued message.

• /handoff to a fresh session. Folds the conversation into a summary in a
  new, light session held by the same agent; asks first whether to keep the old
  session's history or delete it.

• The SESSION card shows what the session costs the machine: the memory and
  CPU of enx and the processes it started (MCP servers, language servers, the
  preview browser), and the history on disk.

• A sub-agent's transcript is read only. Viewing a delegation shows no
  composer; nothing is typed or sent from it.

• The wheel scrolls, a row at a time, and never moves a selection. In
  any list (models, MCP, skills, sessions, the command list, a settings form)
  the wheel scrolls the view and leaves the selection where it is; the next
  key brings the view back to it. The side card scrolls by lines instead of
  turning pages, its position shown as ▲ 21-40/69 ▼. In Settings the wheel
  over a section scrolls it at once, without Enter.

• Every shortcut works on Linux, macOS and Windows. Each one has a form
  that reaches enx through the terminals that swallow the first:

  • Ctrl+S sends now where a terminal cannot tell Ctrl+Enter from Enter;
    Shift+Enter is a newline where it can.
  • Terminals that speak the kitty keyboard protocol (kitty, WezTerm, foot,
    Ghostty, Alacritty, iTerm2) are asked to use it, so those keys are told
    apart there.
  • Alt+Up opens the message menu (macOS takes Ctrl+Up), Alt+V
    attaches from the clipboard (Windows Terminal takes Ctrl+V), and
    Alt+B/Alt+F page the sidebar (macOS sends them for Option+arrows).
  • AltGr characters (@, {, € on many European layouts) type on
    Windows instead of being read as Ctrl+Alt shortcuts.
  • A multi-line paste on Windows arrives as one paste instead of sending
    each line as its own message.
  • Ctrl+Backspace and Alt+Backspace erase a word; Ctrl+D quits only
    from an empty composer and erases forward otherwise; an Alt chord no
    longer types its letter.

• Ctrl shortcuts instead of function keys, since not every terminal passes
  F-keys through: Ctrl+T next sidebar tab, Ctrl+G/Ctrl+X the log's filter
  and detail, and in /model Ctrl+N add, Ctrl+E edit, Ctrl+R refresh.

Models

• Model names are matched however the upstream spells them. claude-opus-4.7,
  claude-opus-4.7-1m, anthropic/claude-opus-4.7:thinking,
  us.anthropic.claude-opus-4-7 and the like now resolve to the right
  catalogue entry, so context window, prices, vision and thinking efforts are
  correct. The maker's listing wins over a reseller's.
• Edit a model's properties (context window, thinking effort, vision,
  prices) from the model list.

Built-in MCP servers

• coolify, dokploy, vps and rag, served by enx itself (no Node or
  Python). All four are listed in /mcp, off by default; Tab turns one on,
  and c opens its setup form. Credentials can also be set from the CLI
  (enx mcp set coolify --url ... --token ..., enx vps add), which an agent
  can run for you, and the running session reloads MCP the moment they land:
  no restart. enx mcp clear <name> forgets a setup. Tokens and passwords go
  to auth.json; secret fields and environment values are redacted from tool
  output; a VPS host key is pinned on first connection.
• RAG indexes by itself and cuts code along its syntax. The workspace
  is indexed when a session starts, a file an agent writes or edits is
  indexed the moment the tool returns, other changes every half minute, and
  everything right before each search, so agents no longer call index
  first. Indexing runs at low OS priority and holds its full scan while a
  turn runs.
  Chunks follow the syntax tree: a function, type, method or config key is
  whole, with its comments, and carries where it sits and what it defines;
  only an item over about 4,000 characters is split, along its own body.
  Markdown is cut at headings, other text at paragraphs. Code that only
  moved keeps its embedding. Turn auto-indexing off in Settings > RAG or
  with --auto-index off.
• vps signs in the way ssh does. Encrypted key files (with their
  passphrase), PuTTY .ppk keys, ssh-agent (Pageant and the OpenSSH agent on
  Windows), hosts writ...

v0.1.0 (early release)

Early release: the first public build of enx. Expect rough edges and bugs; please report them in Issues.

Install:

• macOS / Linux: curl -fsSL https://enowx.ai/install.sh | sh
• Windows (PowerShell): irm https://enowx.ai/install.ps1 | iex

Each archive has a .sha256 checksum beside it.