Skip to content

v0.80.91

Choose a tag to compare

@github-actions github-actions released this 12 Sep 13:59
· 2728 commits to main since this release

0.80.91 introduces Your Office — a society of named AI coworkers with huddles and errands running inside the engine — and A-Coder Cloud v0, a self-hosted always-on agent fleet. The TUI gets a major polish pass: uniform tool-result cards, a consolidated background-task rail, and a /model picker that live-refreshes every provider's catalog. New commands (/context, /diff, /thinking, /memory), plan-mode upgrades, and a fix for keyless local providers round out the release.

Highlights

  • Your Office — named coworkers with faces, huddles (group conversations), and errands (scheduled jobs with memory) persist under ~/.a-coder/cli/office/, each keeping a canonical per-project session so memory and history survive restarts. A user send into a huddle triggers a bounded round-robin drive (max 3 rounds / 10 messages, deterministic @mentions, "(pass)" as silence); holds ("stop @atlas"), supervised approvals routed as office prompts (office_respond), and errands on every/daily/once schedules that run in the coworker's own session. Surfaces: new office_* RPC commands + office_update/office_huddle events (the desktop Office panel) and the /office TUI command (hire, huddle, tell, say, stop, errands); side runtimes via AgentSessionRuntime.createSideRuntime never count against background-session caps.
  • A-Coder Cloud (Sovereign Edition) v0 — self-hosted always-on agent fleet built on Open-PC's architecture (new @earendil-works/pi-cloud package): a-coder cloud spawn <repo> <prompt...> [--base --model --timeout --push] drives an a-coder RPC agent in a cloned workspace — git sync (ac-cloud/<task-id> branches, WIP checkpoint commits, diff summaries, optional origin push), checkpointing, wall-clock budget, extension-UI prompts auto-declined for headless runs — with the agent writing report.md into the repo and the control plane adding a structured report (commits, diffstat, changed files, token/tool usage, warnings) plus session continuity. CLI: spawn, status [--json], review <id>, stop <id>, and the a-coder cloud serve daemon (unix-socket IPC, auto-started on demand, orphaned tasks re-queued after restart).
  • Uniform tool-result cards — settled tool calls render as three-layer cards: an outcome headline with tool-specific stats (40 lines for read, 6 matches in 2 files for grep, +12 −4 for edit, exit 0 · 1.2s for bash), a 3-line content preview with one unified ... N more <unit>, ctrl+shift+o to expand footer, and errors that always break through. MCP and renderer-less extension tools collapse from raw output dumps to the same card (args JSON only when expanded); compact-classified resource reads (AGENTS.md, SKILL.md, docs) stay headline-only.
  • Consolidated TUI zones — background sub-agents and processes render as one status rail flush below the editor (⚡ 2 running · backend (5 tools, 8s) · npm run dev (12s) with a dim ↓ tasks hint), replacing the three scattered surfaces; errors and warnings render as a grouped notice block pinned at the end of the transcript (theme colors, identical notices collapsed with a ×N count, 6 most recent shown) instead of naked Error: lines.
  • Live /model catalog — the picker paints the cached list instantly (never blocking on the network), then re-fetches every fetchable provider — Ollama Cloud, OpenAdapter, LM Studio, llama.cpp, local Ollama — bypassing their 2–5 minute TTL caches, and repaints with newly added models, re-applying your search query so typing is never clobbered; model cycling keeps its TTL cache so rapid cycling stays instant.
  • Keyless local providers fixed — local Ollama, LM Studio, and llama.cpp models no longer error with "No API key found" at selection or prompt time; keyless local providers are treated as configured everywhere, and real stored keys (auth.json or models.json) still take precedence.

Details

Office

  • Office turn_end activity events now carry the coworker turn's stop reason and a willRetry flag, so clients can chime on natural turn ends only — aborts (office Stop), in-session errors and retryable failures stay quiet.
  • While a coworker's turn runs, the engine pushes office_activity RPC events (turn lifecycle, tool start/end with tool name, completed assistant speech — never streaming deltas) so clients can render live office scenes; failed turns surface as a system line in the huddle log with the reason instead of reading as a silent pass.

Engine

  • Built-in read, bash, edit, and write tools prefer strict JSON-schema sampling by default (constrainedSampling: { type: "json_schema", strict: "prefer" }): providers that support strict tools enforce the schema, others fall back silently. Extensions can re-register tool definitions with constrainedSampling: false.
  • Compaction, branch-summarization, and retry spinners render inside the editor top border alongside the working indicator; @earendil-works/pi-tui editor borders gained centered scroll indicators via overridable renderTopBorder/renderBottomBorder hooks, and custom editors opt in with { embedWorkingStatus: true }.
  • SessionManager.inMemory(cwd, options, entries) restores sessions held outside the filesystem (e.g. a database): entries are adopted with their tree, labels and compaction structure intact, a stored session header carries its identity and version through migration, and headerless entries become the body of a fresh session (upstream pi earendil-works#8980 parity).
  • New ui_prompt_start / ui_prompt_end extension events let hosts distinguish active agent work from time spent waiting on blocking ctx.ui prompts (select/confirm/input/editor/custom), with nested prompts tracked by depth; RPC/desktop clients receive them as extension events for waiting-state rendering (upstream pi earendil-works#8355 parity).
  • New pi auth command surface: auth print-api-key, auth print-bearer-token --min-expiry <duration>, and auth check [--json] [--credentials] [--no-refresh] resolve credentials through the model registry and auth storage without starting a session. --no-refresh only refreshes already-expired tokens; --min-expiry requires validity for the requested window and refreshes proactively.
  • ctx.modelRegistry.stream() and streamSimple() for extension model calls through configured providers with resolved authentication; setup failures surface as error events on the returned stream (upstream pi earendil-works#8964/earendil-works#9272 parity).
  • RPC clear_queue retrieves and removes queued steering and follow-up messages for interactive Esc behavior — clear before abort, then restore the returned text in the client editor (upstream pi earendil-works#8432 parity).
  • In-place managed-install updates stage the release tarball, verify its SHA256, and atomically activate it.
  • task_create accepts a tasks array (each with subject, description, optional activeForm/metadata) and creates the complete task list in one call instead of one round-trip per task; ids are assigned sequentially.
  • Full-turn auto-retry honors provider Retry-After: error assistant messages can carry a retryAfterMs hint (extracted from Retry-After/-Ms headers or SDK fields), and the retry waits that long instead of the backoff schedule, capped by the new retry.maxDelayMs setting (default 60000ms).
  • Auto-retry hardening: a turn whose failed attempt already streamed visible output (non-empty text/thinking or any tool call) is no longer silently replayed — the error surfaces instead; and consecutive overloaded/529 failures stop auto-retrying after 3 attempts (reset by any successful turn or a non-overload retry).
  • Grep falls back to grep(1) when ripgrep cannot be provisioned (offline mode, Termux, failed download): the search runs via grep -RInE (with -F/-i/--include/-C mapping) and returns the same path:line: text shape, with a notice that .gitignore rules are not applied. Exposed as runGrepFallback/buildGrepFallbackArguments for tests and remote backends.
  • Teammate mail is no longer spawn-only: a running named teammate drains its inbox after each turn chain and, when mail arrived mid-run, wakes for another turn to act on it (turn budget still respected; mail arriving after the budget is spent stays unread for inspection). Mailbox writes/drains and team-file member updates are serialized through a per-file keyed in-process mutex.

Context & memory

  • AGENTS.override.md per-directory context overrides: when a directory contains one, it is loaded instead of that directory's AGENTS.md/CLAUDE.md; other directories layer normally (upstream pi earendil-works#7681 parity).
  • Persistent-memory auto-injection (opt out via settings.memory.enabled): global and workspace MEMORY.md contents fold into the system prompt as a <persistent_memory> section on prompt (re)build, capped per scope (default 2,000 chars, settings.memory.maxCharsPerScope).
  • /memory picker lists global MEMORY.md, project (workspace) MEMORY.md and the current session's memory file with "(new)" markers for missing files; Enter opens the selected file in your external editor (new files get a minimal # Memory template), and the agent reads the file fresh on its next turn.
  • /rewind now survives a session resume: file-history snapshots are mirrored into the session transcript as file_history_snapshot entries (invisible in LLM context; hidden in the default tree view, labeled checkpoints in "all" view), restored on load and re-hydrated when navigating session-tree branches, so rewind targets the active branch's checkpoint chain.

Commands & plan mode

  • /context visualizes the context window split (System prompt / Tool definitions / Conversation history / Free space) as proportional bars with token counts and a low-space hint.
  • /diff [n] shows the git working-tree status, shortstat and a budgeted colored patch, plus the file-history section: what rewinding the last n user turns would change.
  • /thinking (no argument) opens a searchable thinking-level selector marking the current level and the startup default; Ctrl+S (new app.thinking.save keybinding) saves the selection as the startup default, and a level argument sets it for the session only. AgentSession.setThinkingLevel/cycleThinkingLevel gained { persist } options — session changes no longer rewrite the saved default implicitly.
  • /think [level] (with argument completion) sits next to the existing keybinding cycling, and thinking-keyword escalation is applied at the session layer: "ultrathink" raises to the model's highest supported level, "megathink"/"think harder" to high, "think hard" to medium (raise-only, never lowers, reasoning models only); works in interactive, print, and SDK prompts.
  • Plan-mode enrichment: entering plan mode surfaces a persisted session plan file (<agentDir>/plans/<sessionId>.md) that the model can write while planning (the only write auto-approved in plan mode); obvious read-only bash commands (pwd/ls/cat/head/tail/tree/wc/stat/du/df/which/rg/grep/find/git status|diff|log|show, guarded against redirection, &&/|| chaining, and destructive flags like -delete/-exec) run without an approval prompt; and plan_mode exit accepts plan (content persisted to the plan file) and allowedPrompts (pre-approved commands, e.g. npm test), which become arg-scoped session allow rules taking precedence over the permission mode.
  • Exit-plan approval dialog: when the model leaves plan mode, an interactive dialog previews the plan file and offers "Yes, proceed", "Yes, switch to ask mode" (every edit still prompts), or "No, keep planning" with typed feedback returned to the model while plan mode stays on. Headless/RPC sessions without the handler exit directly, as before.

TUI

  • Background bash processes now notify the agent when they terminate (done/error — by exit or timeout; a user kill from the running-tasks viewer stays quiet): the same <task-notification> wake used for detached sub-agents fires a turn reporting the outcome, including exit code, duration, line/byte totals, the output tail and the full-output log path. Multiple terminations within one wake window batch into a single turn, and a persistently failing wake prompt backs off 2s instead of spinning.
  • Background-task notifications render as compact status pills instead of raw multi-line notes: a completed/failed/killed note collapses to one colored line (glyph + command/subagent summary + duration); the model still receives the full note text.
  • Bash tool cards gain intent tags: [Git] $ git status, [Test·npm test] $ npm test, [Build] $ tsc -p ., [Search·"pattern"] $ rg …, [List] $ ls src — the full command stays visible for approval transparency.
  • Tool cards show a "waiting for approval" phase while their permission prompt is open: the session emits tool_permission_update events around the user prompt, and the pending card renders a warning badge until the call executes.
  • Historical thinking folds away: only the latest round's thinking trace renders in the live transcript; the verbose transcript (Ctrl+Shift+T) temporarily unfolds everything, and the global hide-thinking setting still takes precedence.
  • ask_user_question gains an implicit "Other" row after the model's options that opens a free-text input — type a message and press Enter to send it as the answer (single-select advances, multi-select adds it to the staged selection); Esc returns to the options. The tool description now tells the model the UI provides Other automatically.
  • The running-tasks viewer's agent timeline is fully scrollable: ↑/↓ page back through the complete event history (3 events per tick) with a hidden-events count and a "(↓ returns to live)" affordance, instead of a fixed 12-event tail.
  • Session picker 1-9 quick keys: with an empty search field, pressing 1-9 jumps straight to that row and resumes it.
  • statusLine setting: set statusLine in settings.json to a shell command and its first stdout line renders as an extra row above the footer — Claude-Code-style custom footers without writing an extension. The command runs on a 300ms trailing debounce on session events with a 2s timeout, receiving JSON context on stdin (model, cwd, permissionMode, contextPercent, tokens).
  • Reduced-motion setting (terminal.reducedMotion, env A_CODER_CLI_REDUCED_MOTION=1, toggle in /settings): spinner frames render a static frame, and the message shimmer sweep and in-flight tool blink dots go solid.
  • Background-process output notifications are throttled to ~10Hz (leading + trailing edge) in the store itself: a chatty background command no longer drives per-chunk renders and per-chunk background_processes_update RPC events to the desktop; completion events still notify immediately.
  • Nested slash-command suggestions: typing /office suggests the verbs the command consumes (hire/huddle/tell/say/stop/errands) with usage descriptions, and /export suggests the output formats; both filter as you type.

MCP, extensions & skills

  • The MCP client now negotiates the roots capability and reports the session's working directory to roots/list (root defaults to cwd; resource-loader passes the session cwd). Roots-aware servers like chrome-devtools-mcp scope their file-writing tools to the project instead of falling back to the OS temp directory, so take_screenshot with a filePath inside the session's project works; paths outside stay denied.
  • Skill allowed-tools and paths frontmatter: allowed-tools: Bash(git commit *), Read (CSV string or list) grants those entries as arg-scoped session allow rules when the skill is invoked — by the model via the skill tool or by the user via /skill:name — matching the permission-policy rule syntax. paths: (gitignore-style patterns; glob values starting with * must be YAML-quoted) makes a skill conditional: it stays out of the model's skill listing until a read/write/edit/find call this session touches a matching path, after which it joins the listing on the next prompt. Conditional skills stay invocable by name throughout.
  • Upgraded Composio SDK to @composio/core 0.18.0 + @composio/experimental 0.2.3 (matching pair), silencing the SDK's runtime "Upgrade available" banner.

Analytics

  • Opt-in product analytics via PostHog: when enableAnalytics is on in settings.json (asked during the experimental first-time setup; also toggleable in the desktop Settings panel), the CLI captures anonymous cli_session_start / cli_session_end events (version, mode, OS/arch, model/provider, aggregate token/cost/tool-call counters — never prompt content, paths, or personal data) with the trackingId UUID as distinct ID, plus cli_model_usage — one event per (provider, serving model) pair actually used, so mid-session model switches are counted accurately (requested_model records the picked model when the provider serves a different concrete one, e.g. OpenRouter auto-routing) — and a cli_install event. Events fire in interactive, print, RPC (desktop engine), and ACP modes. Gate overrides: A_CODER_CLI_ANALYTICS=1/0; offline mode always disables. The ingest key is baked into core/analytics.ts with A_CODER_CLI_POSTHOG_KEY/A_CODER_CLI_POSTHOG_HOST overrides for self-hosting/testing.

Fixed

  • Fixed /reload not picking up MCP server changes: the inline extension factory list (including createMcpExtensionFactory) was assembled once at ResourceLoader construction; reload() now rebuilds the inline factories from fresh settings, with previous connections closed cleanly via the session_shutdown (reason reload) event.
  • Fixed direct RPC steer and follow_up commands bypassing extension input handlers (#8718): both now run through the same input-handler/queue path as interactive submissions, with source: "rpc".
  • Fixed RPC abort reporting success without cancelling an in-progress manual compaction (#8920): abort now also cancels compaction and branch summarization, and the session exposes isIdle/waitForIdle() so abort settles before responding.
  • Fixed large tool results crossing the auto-compaction threshold being sent to the provider before compaction: threshold compaction now runs between tool execution and the next assistant response in the same run (#6879, upstream earendil-works#8782). prepareNextTurn now runs only when a next turn actually happens, after shouldStopAfterTurn.
  • Fixed extension messages sent with triggerTurn: false while the agent is running being inserted between a tool call and its result, which made providers that validate message order reject the replayed history; they are now appended once the turn's tool results are in (#8537).
  • Fixed compaction and branch summaries forcing toolChoice: "none", which OpenAI-compatible gateways reject when no tools are sent (#8649, #8638).
  • Fixed read, write, edit, grep, find, ls, and bash tools ignoring ctx.cwd when registered by extensions with a different session cwd (#8627).
  • Fixed proxied plain-HTTP provider requests hanging after a tool call by tunneling them with CONNECT (proxyTunnel, upstream earendil-works#8134).