atris 3.56.0 → 3.56.2

package/atris/skills/engines/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: engines
 description: "Dispatch work to an installed terminal agent or named Atris engine profile. Supports Atris Fast, Claude, Codex, Cursor, Fable, Composer, Haiku, Devin, Grok, and Antigravity (agy). Triggers on: use codex, use cursor, use devin, use grok, use agy, use antigravity, use fable, use claude, use atris, engine, dispatch to, worker agent, second opinion build."
-version: 1.4.0
+version: 1.4.1
 tags:
   - engines
   - claude
@@ -28,6 +28,13 @@ One contract, ten live profiles. The orchestrator writes a bounded task prompt,
 - build: `atris engine <name> <task-id>`
 - switch: `atris engine <name>`
 
+## Fable means Fable
+
+Use `atris engine fable "<bounded question>"` for Fable's take. The canonical
+profile pins `claude-fable-5`, gives deep Fable asks ten minutes by default,
+and records the resolved model in the receipt. An explicit `--model` or
+`--timeout` still wins.
+
 ## Raw binary fallback and debugging
 
 Raw spawns are not the default because they skip Atris receipts, watch, and coaching.
@@ -38,7 +45,7 @@ Raw spawns are not the default because they skip Atris receipts, watch, and coac
 | Claude | `claude -p "<prompt>"` | Uses the local Claude configuration. Add `--model opus` for maximum-depth review or `--model sonnet` for speed. |
 | Codex | `codex exec --dangerously-bypass-approvals-and-sandbox -o <result-file> "<prompt>"` (run from the target repo/worktree; read-only research: `--sandbox read-only`) | Headless, exits when done — run it as a tracked background Bash task like the other engines and completion auto-wakes the session. Final answer lands in the `-o` file. Verified live 2026-08-11. Outside a git repo add `--skip-git-repo-check` or it exits 1. ONE-SHOT SESSIONS (`claude -p` workers): run codex in the FOREGROUND and wait — a one-shot session never wakes again, so backgrounding strands the result (observed 2026-08-11). The old plugin path (`codex-companion.mjs task --background`) is deprecated for dispatch: its job store never notifies the session (a finished result sat unread overnight, 2026-08-10) |
 | Cursor | `cursor-agent --trust -p "<prompt>"` (run from the target repo) | Headless print mode; `--trust` required for non-interactive |
-| Fable | `claude -p "<prompt>" --model opus` | Maximum-depth Claude review. The `fable` Atris profile is the leader lane and uses the installed Claude CLI. |
+| Fable | `claude -p "<prompt>" --model claude-fable-5` | Raw fallback only. Prefer `atris engine fable` so the model, longer timeout, live log, and receipt stay honest. |
 | Composer | `atris run "<objective>" --engine composer` | Fast navigator/executor profile routed through the installed `ax` binary. |
 | Haiku | `claude -p "<prompt>" --model claude-haiku-4-5` | Fast validation and bounded read-only checks. |
 | Devin | `devin -p --permission-mode dangerous -- "<prompt>"` (run from the target repo) | Default permission mode is read-only for writes — build work NEEDS `--permission-mode dangerous`, so only run it in an isolated worktree. Also `devin cloud` for sessions that outlive this machine. Supports `--model swe-1.7` |
@@ -65,7 +72,8 @@ Each engine CLI can pin a specific model. Current best picks:
 
 | Engine | Flag | Best models today |
 |--------|------|-------------------|
-| Claude / Fable | `--model opus` | `opus` currently resolves to Opus 5; use the explicit Opus 4.8 identifier only for reproducibility |
+| Claude | `--model opus` | `opus` follows the local Claude alias; use an explicit model identifier for reproducibility |
+| Fable | `--model claude-fable-5` | Canonical Fable 5 route; `atris engine fable` pins it automatically |
 | Devin | `--model swe-1.7` | `swe-1.7` (free right now: use it as the volume executor for parallel bounded slices), `swe-1.7-lightning` for speed |
 | Cursor | `--model cursor-grok-4.6-xhigh` | `cursor-grok-4.6-xhigh` for second-opinion builds, `cursor-grok-4.6-high-fast` for quick pinned asks (answered in ~12s live 2026-08-12), `composer-2.5` for fast edits; parameterized Claude via `'claude-opus-4-8[effort=high]'`; `--list-models` shows the full menu |
 | Composer | `--engine composer` | `composer 2.5` through the Atris profile |

package/atris/skills/x-search/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: x-search
 description: "X/Twitter search via xAI Grok API. Use when user wants to search tweets, monitor topics, find viral posts, or run social listening. Costs 5 credits per search. Triggers on x search, tweet search, twitter search, social listening, revenue intel, viral tweets."
-version: 2.0.0
+version: 2.2.0
 tags:
   - x-search
   - social
@@ -12,6 +12,25 @@ tags:
 
 > Drop this in `~/.claude/skills/x-search/SKILL.md` and Claude Code becomes your X/Twitter intelligence tool.
 
+## Feature map
+
+Hard path. Follow in order. Soft memory does not override this.
+
+```
+login
+    |
+   x-search QUERY (5 credits)
+    |
+   write one Apply (claimable)
+    |
+   stop
+```
+
+1. Login: stored login, then `atris login --agent` if needed. Never paste tokens. Never `/auth/cli`.
+2. Search: `atris x-search "QUERY"`. Always billed (5 credits). Empty or failed search refunds.
+3. One Apply: name a product change plus a receipt that can fail. If missing, the CLI writes a fill-this stub and a next-line. Search still exits 0.
+4. Stop. Empty or failed search does not owe an Apply.
+
 ## Customer path (preferred)
 
 Logged-in customers should use the CLI. Same auth and billing as `atris youtube process` (5 credits).

package/atris/skills/youtube/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: youtube
-description: "YouTube discovery and learning. Use atris youtube search QUERY to get free youtu.be links (local ytsearch/yt-dlp, zero credits). Use atris youtube search --paid QUERY to buy watch permalinks from Atris (5 credits, login required). A YouTube link in any message routes here: run atris youtube notes URL FIRST (free, about 30s, quotes verified). Never summarize a video from model memory. Use atris youtube process only to store it as queryable knowledge (5 credits). Triggers on: youtube search, find videos, paid youtube search, any youtube.com or youtu.be link, youtube, video, watch this, notes on this."
-version: 2.7.0
+description: "YouTube discovery and learning. Get watch permalinks with atris youtube search QUERY (free, local ytsearch/yt-dlp). On 429 the CLI already retries; use cached rows if printed, else STOP. Never run --paid after a 429. --paid only when the user explicitly asked to buy permalinks (5 credits). After a URL is picked, atris youtube notes URL (free, ephemeral unless --save). atris youtube process only to store knowledge (5 credits). Never paste tokens. Never /auth/cli. Mint with atris login --agent from a stored login. Never summarize a video from model memory. Triggers on: youtube search, find videos, paid youtube search, any youtube.com or youtu.be link, youtube, video, watch this, notes on this."
+version: 2.11.0
 tags:
   - youtube
   - research
@@ -11,47 +11,56 @@ tags:
 
 # YouTube Skill
 
-Three rails. Pick before running anything:
+## Feature map
 
-- **Search / discovery (free)** → `atris youtube search "<query>"`. Local ytsearch or yt-dlp `ytsearchN:`. Returns title, channel, duration, views, upload_date, and a `youtu.be` link. Zero credits. Use this to *get* video links; do not call process until the user picks one.
-- **Paid search (5 credits)** → `atris youtube search --paid "<query>"`. Posts `/youtube/search` with the stored token or a youtube-scope agent mint. Prints watch permalinks, titles, and credits. Use this when the customer wants to buy permalinks the same way as `atris x-search`. Do not replace free search with this.
-- **Notes / learning (free)** → `atris youtube notes <url>`. Local captions + a fast engine, about 30 seconds, quotes verified against the transcript. Use when the goal is to learn from a video you already have.
-- **Process / product (5 credits)** → `atris youtube process <url>`. Credits-billed cloud knowledge store. Use when a customer/agent needs the video stored as Atris knowledge.
+Hard path. Follow in order. Soft memory does not override this.
 
-If the user says "find videos", "search youtube", or "get youtube links" → search. If they say "learn from", "notes on", "alpha", or "rabbit hole" → notes. If they say "process", "store", "add to knowledge" → process.
+```
+search QUERY (free)
+    |
+   429 --> CLI already retried once
+        --> cached rows printed? use them
+        --> rate-limit sentence printed? STOP
+             do not run --paid
+    |
+   pick a URL --> notes URL (free, ephemeral unless --save)
+    |                 --save files the brief + apply stub
+    |            or teach URL [--section N] (one chapter: claim numbers, named mechanisms, one check; free unless --save)
+    |
+   write one Apply (claimable) before process
+    |
+   store knowledge? --> process URL (5 credits)
+
+--paid QUERY only if the user asked to buy permalinks
+login: atris login --agent from a stored login
+never paste tokens, never /auth/cli
+```
+
+1. Get watch permalinks: `atris youtube search QUERY` (free).
+2. If 429: wait/retry is already in the CLI. If it prints cached rows, use those. If it prints `youtube rate-limited local search. do not use --paid as a fallback; retry later.`, STOP. Do not run `--paid`.
+3. `--paid` only when the user explicitly asked to buy permalinks. The CLI hard-refuses `--paid` when the free cache still has a fresh same-query hit.
+4. `atris youtube notes URL` after a URL is picked (free). Notes is ephemeral unless `--save`.
+5. Write one Apply (change + receipt) before `atris youtube process`. Process still requires a filled Apply (so you `--save`, fill Apply, then process).
+6. Never paste tokens. Never `/auth/cli`. Mint with `atris login --agent` from a stored login.
+
+If the user says "find videos", "search youtube", or "get youtube links" → search. If they say "learn from", "notes on", "alpha", or "rabbit hole" → notes. If they say "process", "store", "add to knowledge" → process. "Buy permalinks" or "paid search" is the only ask that unlocks `--paid`.
 
 Never summarize a video from model memory; that is fabrication.
 
-## Bootstrap (ALWAYS Run First)
+## Bootstrap
+
+Search and notes need no login. Paid search and process need a stored login, then `atris login --agent`. Never print credentials. Never `/auth/cli`.
 
 ```bash
 #!/bin/bash
 set -e
 
-# 1. Check atris CLI
 if ! command -v atris &> /dev/null; then
   echo "Installing atris CLI..."
   npm install -g atris
 fi
 
-# 2. Optional login (required only for process / credits)
-if [ ! -f ~/.atris/credentials.json ]; then
-  echo "Not logged in. Search and notes still work. Process needs: atris login"
-fi
-
-# 3. Extract token when present
-if [ -f ~/.atris/credentials.json ]; then
-  if command -v node &> /dev/null; then
-    TOKEN=$(node -e "console.log(require('$HOME/.atris/credentials.json').token)")
-  elif command -v python3 &> /dev/null; then
-    TOKEN=$(python3 -c "import json,os; print(json.load(open(os.path.expanduser('~/.atris/credentials.json')))['token'])")
-  elif command -v jq &> /dev/null; then
-    TOKEN=$(jq -r '.token' ~/.atris/credentials.json)
-  fi
-  export ATRIS_TOKEN="$TOKEN"
-fi
-
-echo "Ready. YouTube skill active (search + notes free; process 5 credits)."
+echo "Ready. Feature map: free search, 429 cache-or-stop, notes, then process."
 ```
 
 ---
@@ -66,14 +75,18 @@ atris youtube search "MCP agents" --json
 
 Uses `ytsearch` on PATH when present, else bundled `scripts/det/ytsearch`, else `yt-dlp --flat-playlist --print` with `ytsearchN:`. No credits. No `/agent/process_youtube` call.
 
-## Paid search (5 credits)
+On 429 the CLI retries once, then serves `~/.atris/youtube-search-cache.json` if the same query is younger than one hour. If it prints the rate-limit sentence, stop. Do not run `--paid`.
+
+## Paid search (5 credits, opt-in buy only)
+
+Only when the user explicitly asked to buy permalinks. A 429 or a missing cache is not that ask.
 
 ```bash
 atris youtube search --paid "MCP agents 2026"
 atris youtube search --paid "MCP agents" --limit 10
 ```
 
-Requires login. Uses the stored token, or mints a youtube-scope agent token from disk the same way as `atris youtube process` and `atris x-search`. Never `/auth/cli`. Prints `title | watch permalink` plus credits. Empty or failed searches refund.
+Requires a stored login, then `atris login --agent`. The CLI mints a youtube-scope agent token from disk the same way as `atris youtube process` and `atris x-search`. Never `/auth/cli`. Never paste tokens. Prints `title | watch permalink` plus credits. Empty or failed searches refund.
 
 Line contract:
 
@@ -90,10 +103,7 @@ title | channel | duration | views | upload_date | https://youtu.be/ID
 Base: `https://api.atris.ai/api`
 Auth: `-H "Authorization: Bearer $TOKEN"`
 
-### Get Token
-```bash
-TOKEN=$(node -e "console.log(require('$HOME/.atris/credentials.json').token)")
-```
+Do not extract or paste tokens. The CLI reads the stored login. Agents mint with `atris login --agent`.
 
 ### Process a Video
 ```bash
@@ -203,7 +213,7 @@ Two layers, never mixed. The reply the person reads is flowing prose: ideas, spe
 
 `atris youtube search` shells to local ytsearch/yt-dlp and never hits the Atris API.
 
-`atris youtube search --paid` posts `{query, limit}` to `/youtube/search` with bearer auth. Agent tokens need the youtube scope.
+`atris youtube search --paid` posts `{query, limit}` to `/youtube/search` with bearer auth. Agent tokens need the youtube scope. This is an opt-in buy, not a 429 fallback.
 
 `atris youtube` process first tries local transcript extraction with `yt-dlp`. It sends timestamped `transcript_text` to `/agent/process_youtube` with `cache_transcript=false`. If local transcript processing fails with a retryable error, it falls back to cloud video processing. Use `--json` to inspect `metadata.processing_method` and `metadata.transcript_source`.
 
@@ -225,35 +235,33 @@ Two layers, never mixed. The reply the person reads is flowing prose: ideas, spe
 
 | Error | Meaning | Fix |
 |-------|---------|-----|
-| `401` | Token expired/invalid | `atris login --force` |
+| `401` | Token expired/invalid | `atris login --agent` from a stored login |
 | `402` | Not enough credits | Check balance, purchase at atris.ai |
 | `400` | Invalid YouTube URL | Check URL format |
 | `502` | Transcript or cloud processing failed | Retry; credits auto-refunded when backend fails |
 | search exit 2 | ytsearch/yt-dlp missing or no results | Install yt-dlp, or put ytsearch on PATH |
-| search 429 | YouTube rate-limited local search | serve last same-query free rows if younger than one hour; otherwise retry later; do not use --paid |
+| search 429 | YouTube rate-limited local search | CLI already retried; use cached rows if printed; if the rate-limit sentence prints, STOP; do not use --paid |
 
 ---
 
 ## Quick Reference
 
 ```bash
-# Setup (once)
+# Setup: human stores a login, then agents mint. never /auth/cli
 npm install -g atris && atris login
+atris login --agent
 
 # Free: discover videos
 atris youtube search "MCP agents 2026"
 atris youtube search "MCP agents" --limit 10
 
-# Paid: watch permalinks from Atris (5 credits)
+# Paid: only if the user asked to buy permalinks (5 credits). never a 429 fallback
 atris youtube search --paid "MCP agents 2026"
 
 # Free: notes on a URL you already have
 atris youtube notes "https://youtu.be/VIDEO_ID"
 
-# Get token (process only)
-TOKEN=$(node -e "console.log(require('$HOME/.atris/credentials.json').token)")
-
-# Process a video (5 credits)
+# Process a video (5 credits). different store from paid search
 atris youtube process "https://youtube.com/watch?v=..." --query "Create a outline (flowing, idea-first) and action brief"
 
 # Process + store to agent knowledge

package/bin/atris.js

@@ -47,7 +47,7 @@ const {
   requireAccountBound,
   refuseAccountGlobal,
 } = require('../lib/account-bound');
-const { isHelpToken, argsWantHelp } = require('../lib/noninteractive');
+const { isHelpToken, argsWantHelp, hasYesFlag } = require('../lib/noninteractive');
 
 // State detection for smart default
 const { detectWorkspaceState, loadContext } = require('../lib/state-detection');
@@ -56,12 +56,20 @@ const {
   createStarterTask,
   shouldGatherContext,
   isAtrisMetaQuestion,
+  isFlagLikeAnswer,
   renderPrompt: renderContextGathererPrompt,
 } = require('../lib/context-gatherer');
 const {
   buildFirstMinute,
+  firstTalkJson,
+  folderName,
   freshMinuteJson,
   isBareAtrisFlag,
+  isFirstTalkLine,
+  isLeftoverVerbLook,
+  listUserVisibleWork,
+  personName,
+  renderFirstTalk,
   shouldAutoInitFresh,
 } = require('../lib/first-minute');
 
@@ -203,7 +211,7 @@ function parseNaturalEntryArgs(args = []) {
   };
 }
 
-function applyRunnerFlags(args) {
+function applyRunnerFlags(args, options = {}) {
   // --engine <name> is the operator-facing spelling of --runner-profile:
   // one flag rents a specific intelligence for this run.
   const engineFlag = readOptionArg(args, '--engine');
@@ -243,6 +251,9 @@ function applyRunnerFlags(args) {
     process.env.ATRIS_RUNNER_MODEL = runnerModel;
     process.env.ATRIS_CLAUDE_MODEL = runnerModel;
   }
+  // Talk-only and unbound scratch still validate --engine / --runner-profile.
+  // They skip this seed so resolveDefaultEngine cannot mint engines.json.
+  if (options.seed === false) return;
   // No per-run choice and no env: the workspace's saved engine
   // (.atris/engine.json, written by `atris engine <name>`) becomes the
   // profile for every loop spawn. For the loop commands themselves, fall all
@@ -484,17 +495,14 @@ function showStartHelp() {
 }
 
 function showHelpShort() {
-  console.log('atris: you say what you want; atris builds it, checks it, and shows proof.');
-  console.log('');
-  console.log('Golden path:');
-  console.log('  atris init [--yes] [--minimal]   scaffold this project');
-  console.log('  atris                           load context (MAP, tasks, journal)');
-  console.log('  atris task new "..."            file work');
-  console.log('  atris task claim <id> --as <m>  take ownership');
-  console.log('  atris task ready <id> --proof "..."');
-  console.log('  atris review                    validate');
-  console.log('  atris recap                     spoken recap of what landed');
-  console.log('  atris mission status            active goals + next move');
+  console.log('atris: you say what you want. already won. one next step.');
+  console.log('');
+  console.log('  atris "what do you want here?"   empty folder');
+  console.log('  atris later "..."                remember without a room');
+  console.log('  atris do                         work already here');
+  console.log('  atris spaceship                  Keep working for a few hours');
+  console.log('  atris autopilot                  Keep working until you stop');
+  console.log('  atris mission                    the one goal');
   console.log('');
   console.log('More: atris help --all   ·   machine list: atris help --json');
 }
@@ -506,7 +514,7 @@ function helpCatalog() {
     { name: 'review', summary: 'validate work and capture learnings', json: false, interactive_risk: 'low' },
     { name: 'recap', summary: 'spoken recap of what your team did', json: true, interactive_risk: 'low' },
     { name: 'mission', summary: 'durable goal + owner + verifier + tick', json: true, interactive_risk: 'medium' },
-    { name: 'log', summary: 'append an idea to today inbox (or --repl)', json: false, interactive_risk: 'medium' },
+    { name: 'log', summary: 'append an idea to today inbox (or --repl)', json: true, interactive_risk: 'low' },
     { name: 'wish', summary: 'headless inbox / mission intake', json: true, interactive_risk: 'low' },
     { name: 'learn', summary: 'structured project learnings', json: false, interactive_risk: 'low' },
     { name: 'activate', summary: 'load atris context for this session', json: false, interactive_risk: 'low' },
@@ -517,38 +525,16 @@ function helpCatalog() {
 }
 
 function showHelpAll() {
+  console.log('you say what you want. already won. one next step.');
   console.log('');
-  console.log('━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━');
-  console.log('atris');
-  console.log('you say what you want in plain words. atris builds it, checks it, and shows you proof.');
-  console.log('━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━');
-  console.log('');
-  console.log('Quick start:');
-  console.log('');
-  console.log('  1. atris                  load context (MAP, tasks, journal)');
-  console.log('     npx atris              same after a local install');
-  console.log('  2. say what you want built, checked, or researched');
-  console.log('  3. atris files the work, then you review');
-  console.log('');
-  console.log('Common invocations:');
-  console.log('  atris "<request>"         one isolated change, then Review');
-  console.log('  atris "<request>" --verify "<cmd>" --json');
-  console.log('  atris init [--yes]        scaffold this project');
-  console.log('  npx atris init [--yes]    same after a local install');
-  console.log('  atris mission status      current mission + next move');
-  console.log('  atris task ready <id> --proof "..."');
-  console.log('  atris computer');
-  console.log('  atris engine validate latest');
-  console.log('  atris playbook                 workspace style rules and paired trials');
-  console.log('  atris business init "My Company"');
-  console.log('  atris run');
-  console.log('  atris drill');
-  console.log('  atris status');
-  console.log('  atris soul');
-  console.log('  atris fleet status');
-  console.log('  ask/stop/ready            see atris mission and atris task ready');
-  console.log('');
-  console.log('━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━');
+  console.log('  atris later "..."                remember without a room');
+  console.log('  atris do                         work already here');
+  console.log('  atris spaceship                  Keep working for a few hours');
+  console.log('  atris autopilot                  Keep working until you stop');
+  console.log('  atris mission                    the one goal');
+  console.log('  atris recap                      what got done');
+  console.log('  atris review                     check the work');
+  console.log('  atris stop                       stop the live goal');
   console.log('');
   console.log('Atris Computers:');
   console.log('  Owner = User | Business');
@@ -649,6 +635,8 @@ function showHelpAll() {
   console.log('  experiments init [slug]     - Prepare atris/experiments/ or scaffold a pack');
   console.log('  experiments validate        - Validate experiment packs');
   console.log('  experiments run <slug>      - Execute a pack or record an Endstate receipt');
+  console.log('  experiments keep <slug>     - Keep a pack only when measure.py moves 0 to 1');
+  console.log('  experiments revert <slug>   - Run a pack reset.py after a refused keep');
   console.log('  experiments benchmark [m]   - Run validate/runtime experiment benchmarks');
   console.log('  bench      - run core benchmark gates');
   console.log('');
@@ -765,7 +753,7 @@ function showHelpAll() {
   console.log('');
   console.log('Other:');
   console.log('  version    - Show Atris version');
-  console.log('  help       - Short golden path (help --all for every command)');
+  console.log('  help       - you say what you want (help --all for every command)');
   console.log('');
   console.log('Tip: Just run "atris" to get started');
   console.log('');
@@ -1344,13 +1332,13 @@ function printFirstUseNext() {
   console.log(`Then: ${firstUseCommand()}`);
 }
 
-function runMinimalInit() {
+function runMinimalInit({ silent = false } = {}) {
   const script = process.argv[1] || path.join(__dirname, 'atris.js');
   const result = spawnSync(process.execPath, [script, 'init', '--minimal', '--yes'], {
     cwd: process.cwd(),
     env: process.env,
     encoding: 'utf8',
-    stdio: 'inherit',
+    stdio: silent ? ['ignore', 'pipe', 'pipe'] : 'inherit',
   });
   return Number.isInteger(result.status) ? result.status : 1;
 }
@@ -1366,16 +1354,56 @@ async function interactiveEntry(userInput, options = {}) {
   const workspaceDir = process.cwd();
   const state = detectWorkspaceState(workspaceDir);
 
-  // Fresh folder is the first-minute product: one next command, no menu.
+  // Fresh folder is a conversation, not a scaffold form.
   if (state.state === 'fresh') {
-    if (options.asJson) {
-      console.log(JSON.stringify(freshMinuteJson(), null, 2));
-      return 2;
+    const request = String(userInput || '').trim();
+    const room = folderName(workspaceDir);
+    const who = personName();
+    if (!request) {
+      if (options.asJson) {
+        console.log(JSON.stringify(freshMinuteJson(room, listUserVisibleWork(workspaceDir), { root: workspaceDir }), null, 2));
+        return 2;
+      }
+      if (shouldAutoInitFresh(process.argv.slice(2))) {
+        return runMinimalInit();
+      }
+      printFirstMinuteScreen({ root: workspaceDir, fresh: true });
+      return 0;
     }
-    if (shouldAutoInitFresh(process.argv.slice(2))) {
-      return runMinimalInit();
+    if (isLeftoverVerbLook(request)) {
+      if (options.asJson) {
+        console.log(JSON.stringify(freshMinuteJson(room, listUserVisibleWork(workspaceDir), { root: workspaceDir }), null, 2));
+        return 2;
+      }
+      printFirstMinuteScreen({ root: workspaceDir, fresh: true });
+      return 0;
     }
-    printFirstMinuteScreen({ root: workspaceDir, fresh: true });
+    if (!isFirstTalkLine(request) && isAtrisMetaQuestion(request)) {
+      if (options.asJson) {
+        console.log(JSON.stringify({
+          schema: 'atris.overview.v1',
+          ok: true,
+          product: 'Atris',
+        }, null, 2));
+        return 0;
+      }
+      printAtrisOverview();
+      return 0;
+    }
+    if (isFlagLikeAnswer(request)) {
+      printFirstMinuteScreen({ root: workspaceDir, fresh: true });
+      return 0;
+    }
+    const initStatus = runMinimalInit({ silent: true });
+    if (initStatus !== 0) return initStatus;
+    saveContextProfile(workspaceDir, request, { source: 'first_talk' });
+    const starter = createStarterTask(workspaceDir, request);
+    if (options.asJson) {
+      console.log(JSON.stringify(firstTalkJson({ starter, person: who, folder: room }), null, 2));
+      return 0;
+    }
+    console.log('');
+    console.log(renderFirstTalk({ person: who, folder: room, starter }));
     return 0;
   }
 
@@ -1921,18 +1949,14 @@ if (command === 'init') {
     .catch((err) => { console.error(`\n✗ Error: ${err.message || err}`); process.exit(1); });
 } else if (command === 'land') {
   // landCommand can throw synchronously on empty repos; wrap so callers get a
-  // one-line error instead of an uncaught stack from Promise.resolve(fn()).
+  // one-line sentence instead of an uncaught stack from Promise.resolve(fn()).
   Promise.resolve()
     .then(() => require('../commands/land').landCommand(process.argv.slice(3)))
-    .then((code) => process.exit(code || 0))
+    .then((code) => process.exit(typeof code === 'number' ? code : 0))
     .catch((err) => {
-      const msg = String(err && err.message || err || '');
-      if (/no master\/main|no commits|not a git/i.test(msg)) {
-        console.error(msg);
-      } else {
-        console.error(`land: ${msg}`);
-      }
-      process.exit(1);
+      const msg = String(err && err.message || err || 'land could not run').split('\n')[0];
+      console.log(msg);
+      process.exit(2);
     });
 } else if (command === 'caretaker') {
   Promise.resolve(require('../commands/caretaker').caretakerCommand(process.argv.slice(3)))
@@ -2035,7 +2059,6 @@ if (command === 'init') {
 } else if (command === 'log') {
   const logArgs = process.argv.slice(3);
   const subcommand = logArgs[0];
-  const positional = logArgs.filter((arg) => !String(arg).startsWith('-'));
   if (subcommand === 'sync') {
     require('../commands/log-sync').logSyncAtris()
       .then(() => process.exit(0))
@@ -2044,24 +2067,16 @@ if (command === 'init') {
         process.exit(1);
       });
   } else if (subcommand === '--help' || subcommand === '-h' || subcommand === 'help') {
-    console.log('Usage: atris log ["note"] | --repl | sync | <business-slug> | help');
+    console.log('Usage: atris log ["note"] [--json] | --repl | sync | help');
     console.log('');
     console.log('  atris log "an idea"      Append a note to today\'s ## Inbox');
+    console.log('  atris log text           One word still lands in Inbox. No live business required');
+    console.log('  atris log "an idea" --json');
     console.log('  atris log --repl         Open today\'s journal REPL');
     console.log('  atris log                Open the REPL (needs a terminal)');
     console.log('  atris log sync           Sync the local journal to cloud');
-    console.log('  atris log <slug>         Show business log history (single slug token)');
     console.log('  atris wish "..."         Headless inbox write; add --json --no-mission for agents');
     process.exit(0);
-  } else if (
-    positional.length === 1
-    && !logArgs.includes('--repl')
-    && /^[a-z0-9][a-z0-9_-]{0,63}$/i.test(positional[0])
-  ) {
-    // Business log: atris log <business-slug> (single slug token only)
-    require('../commands/context-sync').businessLog(positional[0])
-      .then(() => process.exit(0))
-      .catch((err) => { console.error(`\n✗ Error: ${err.message || err}`); process.exit(1); });
   } else {
     logCmd();
   }
@@ -2084,13 +2099,14 @@ if (command === 'init') {
   }
 } else if (command === 'now') {
   const args = process.argv.slice(3);
-  if (args.includes('--help') || args.includes('-h') || args[0] === 'help') {
-    require('../commands/now').nowAtris(args);
-    process.exit(0);
-  }
-  require('../commands/now').nowAtris(args);
+  const nowCode = require('../commands/now').nowAtris(args);
+  process.exit(Number.isInteger(nowCode) ? nowCode : 0);
 } else if (command === 'recap') {
-  require('../commands/recap').recapAtris(process.argv.slice(3));
+  const recapCode = require('../commands/recap').recapAtris(process.argv.slice(3));
+  if (Number.isInteger(recapCode)) process.exit(recapCode);
+} else if (command === 'later') {
+  const laterCode = require('../commands/later').laterAtris(process.argv.slice(3));
+  process.exit(Number.isInteger(laterCode) ? laterCode : 0);
 } else if (command === 'activate') {
   const args = process.argv.slice(3);
   if (args.includes('--help') || args.includes('-h') || args[0] === 'help') {
@@ -2431,7 +2447,14 @@ if (command === 'init') {
     process.exit(0);
   }
 
-  applyRunnerFlags(args);
+  const { isUnboundScratchFolder } = require('../lib/scratch-root');
+  const { runObjective, runMissionFront } = require('../commands/run-front');
+  const runTalkOnly = !runObjective(args) && !hasYesFlag(args) && !args.includes('--legacy');
+  // Flag validation always runs. The engine-registry seed is what mints
+  // .atris/state/engines.json, so first-talk and unbound scratch skip only that.
+  applyRunnerFlags(args, {
+    seed: !runTalkOnly && !isUnboundScratchFolder(process.cwd()),
+  });
 
   if (args.includes('--legacy')) {
     const legacyArgs = args.filter(a => a !== '--legacy');
@@ -2451,7 +2474,7 @@ if (command === 'init') {
         process.exit(1);
       });
   } else {
-    require('../commands/run-front').runMissionFront(args)
+    runMissionFront(args)
       .then((code) => process.exit(code || 0))
       .catch((error) => {
         console.error(`\u2717 Run failed: ${error.message || error}`);
@@ -2463,7 +2486,21 @@ if (command === 'init') {
   process.exit(code);
 } else if (command === 'autopilot') {
   const args = process.argv.slice(3);
-  applyRunnerFlags(args);
+  const { isUnboundScratchFolder, refuseUnboundScratch } = require('../lib/scratch-root');
+  const { resolveWorkspaceRoot } = require('../lib/mission-root');
+  const autopilotRoot = resolveWorkspaceRoot(process.cwd());
+  const autopilotHelp = argsWantHelp(args) || args[0] === 'help';
+  const autopilotControl = args[0] === 'stop' || args[0] === 'status';
+  const autopilotJsonPlan = args.includes('--json') && !args.includes('--yes') && !args.includes('-y');
+  const autopilotTalkOnly = !hasYesFlag(args) && !args.includes('--auto') && !args.includes('--legacy');
+  // applyRunnerFlags seeds .atris/state/engines.json. Help, stop/status, JSON
+  // plan, and an unbound scratch talk must not mint a room. --yes is not an unlock.
+  if (!autopilotHelp && !autopilotControl && !autopilotJsonPlan && !autopilotTalkOnly && isUnboundScratchFolder(autopilotRoot)) {
+    process.exit(refuseUnboundScratch());
+  }
+  if (!autopilotHelp) {
+    applyRunnerFlags(args, { seed: !isUnboundScratchFolder(autopilotRoot) });
+  }
 
   if (args.includes('--legacy')) {
     const legacyArgs = args.filter(a => a !== '--legacy');
@@ -2739,8 +2776,9 @@ if (command === 'init') {
     const raw = process.argv.slice(3);
     const { gmailCommand, extractGmailMailboxAccount } = require('../commands/integrations');
     const mailbox = extractGmailMailboxAccount(raw);
-    const gateInput = mailbox.mailboxAccountId ? ['--account', ...mailbox.args] : mailbox.args;
-    const gate = requireAccountBound(gateInput);
+    // Mailbox --account <id> is a selector, not the workspace opt-in.
+    // Scratch folders refuse every gmail verb the same way as slack/inbox.
+    const gate = requireAccountBound(mailbox.args.filter((arg) => arg !== '--account'));
     if (!gate.ok) {
       process.exit(refuseAccountGlobal());
       return;
@@ -2775,9 +2813,15 @@ if (command === 'init') {
 } else if (command === 'slack') {
   {
     const raw = process.argv.slice(3);
-    const gate = requireAccountBound(raw);
-    if (!gate.ok) process.exit(refuseAccountGlobal());
-    const { slackCommand } = require('../commands/integrations');
+    const { slackCommand, extractGmailMailboxAccount } = require('../commands/integrations');
+    const account = extractGmailMailboxAccount(raw);
+    // Account --account <id> is a selector, not the workspace opt-in.
+    // Scratch folders refuse every slack verb the same way as gmail/inbox.
+    const gate = requireAccountBound(account.args.filter((arg) => arg !== '--account'));
+    if (!gate.ok) {
+      process.exit(refuseAccountGlobal());
+      return;
+    }
     const subcommand = gate.args[0];
     const args = gate.args.slice(1);
     slackCommand(subcommand, ...args)