fix: v2.7.4 — a developer's System page carries no host or other-user path

- GET /api/settings/info for a developer: no DB path, no hooks file path,
  and no transcript-cache keys — those were EVERY user's transcript paths
  (their folders and session ids). GET /api/updates/status: no repo_root.
  Test: server-mode-leaks.
- Also in this release (earlier commits): a member's missing workspace
  folder is made as them at start; --allow-dangerously-skip-permissions only
  for an account that accepted the bypass warning or a bypass run.
Full suites: server 2889/2889, client 1250/1250, tsc -b.

CLAUDE.md

@@ -1,6 +1,6 @@
 # Claude Code Working Guide
 
-_Last synced: v2.7.4 (2026-10-12) — Changes hides a staged change undone on disk (#72); chat says when the session has not started (#73)._ A change that makes any statement in this file, `AGENTS.md`, or a README wrong — a moved path, a new or changed extension point, a retired rule — updates that file in the same change-set and bumps this marker. **Bump = replace the version/date and a ≤ 1-line summary; never append a changelog here** — details belong in the commit message and `docs/`, because this whole file is loaded into every session.
+_Last synced: v2.7.4 (2026-10-12) — Changes hides a staged change undone on disk (#72); chat says when the session has not started (#73); a member's missing folder is made as them, bypass offered only once accepted._ A change that makes any statement in this file, `AGENTS.md`, or a README wrong — a moved path, a new or changed extension point, a retired rule — updates that file in the same change-set and bumps this marker. **Bump = replace the version/date and a ≤ 1-line summary; never append a changelog here** — details belong in the commit message and `docs/`, because this whole file is loaded into every session.
 
 ## Project mission
 - Maintain a reliable local-first dashboard for Claude Code session monitoring.

docs/API.md

@@ -359,7 +359,7 @@ Error `409 ESESSIONLIVE` is returned when `resumeSessionId` (or `resume`) names
 
 Error `400 EBADPROMPT` is returned for a `prompt` (or `POST /api/run`'s `initialPrompt`) over 128 KB: no CLI can take one argument that long (Linux's `MAX_ARG_STRLEN`, issue #68) — attach the text as a file instead. A prompt over 4 KB never rides in tmux's argv (one tmux message holds 16 KB, `command too long`): outside Windows it is written to `<data dir>/run-prompts/<run id>.txt` (0600) and the pane's own `sh` reads it back as the CLI's last argument and deletes the file. An `after-start` tool's paste still goes through `tmux set-buffer` and stays under that 16 KB.
 
-Error `409 ENOCWD` is returned when the workspace's folder no longer exists (or is not a directory) — the pane would otherwise start in the account's home under this workspace's name. A folder this server cannot read (`EACCES`) is not refused.
+Error `409 ENOCWD` is returned when the workspace's folder no longer exists (or is not a directory) — the pane would otherwise start in the account's home under this workspace's name. A folder this server cannot read (`EACCES`) is not refused. Server mode with one Linux user per account: a missing folder inside a member's own home is created instead, as that member (`mkdir -p` through `write-as`) — their home itself can be no workspace, so this is how a new member's first project folder comes to exist.
 
 Error `409 ERUNLIVE` is returned if the workspace already has a live run; stop it first with the `stop` action. The same code and status come back when the workspace's tmux pane is already running an agent while a `prompt` or `resumeSessionId` was sent — that argv cannot be typed over a live agent, so the request is refused instead of answered 200 with nothing done. A start that asks for neither still adopts the existing pane silently.
 

docs/SERVER-MODE.md

@@ -822,8 +822,20 @@ agenthub users provision alice      # the Linux account only (a user made before
   writes three `auth.log` lines); a pane the dashboard starts or stops shows at
   once, one that exits on its own within those 5 s.
 - Claude Code does not carry a folder's trust down from the member's home
-  (which provisioning trusts): a member answers "Trust this folder" once per
-  workspace, in the terminal.
+  (which provisioning trusts): a member may be asked "Trust this folder" once
+  per workspace, in the terminal. AgentHub never answers it for them — it is
+  what keeps a cloned repository's own settings and hooks from running unasked.
+- A member's first project folder: a workspace named on a folder that does not
+  exist yet inside their home is created at its first start, AS the member
+  (from v2.7.4; their home itself is no workspace).
+- Bypass in a pane's shift+tab cycle (`--allow-dangerously-skip-permissions`)
+  is offered only to an account that already accepted Claude Code's bypass
+  warning (`skipDangerousModePermissionPrompt` in its `~/.claude/settings.json`)
+  or to a run that starts in bypass itself. Before v2.7.4 every new member's
+  first session opened on that warning, which the Chat UI cannot answer.
+- A developer's System page (`GET /api/settings/info`, `GET /api/updates/status`)
+  carries no path: not the database's, the hooks file's, the install folder,
+  nor the transcript cache's keys — those are every user's transcript paths.
 
 ### Undo and uninstall
 

server/__tests__/server-mode-leaks.test.js

@@ -148,3 +148,23 @@ describe("server mode: a member's skills/commands come from their own workspaces
     assert.ok(names(r).includes("max-skill"));
   });
 });
+
+describe("the System page's facts for a developer", () => {
+  // Every string that looks like an absolute path, anywhere in a body.
+  const paths = (o, out = []) => {
+    if (typeof o === "string" && o.startsWith("/")) out.push(o);
+    else if (o && typeof o === "object") for (const v of Object.values(o)) paths(v, out);
+    return out;
+  };
+
+  it("carries no host or other-user path (DB, hooks file, transcript cache keys)", async () => {
+    const mia = await login("mia", "member-pass-1");
+    const info = await call("/api/settings/info", mia);
+    assert.equal(info.status, 200);
+    assert.deepEqual(paths(info.body), []);
+    assert.equal(typeof info.body.db.size, "number", "the numbers stay");
+    const boss = await login("boss", "admin-pass-1");
+    const full = await call("/api/settings/info", boss);
+    assert.ok(full.body.db.path, "an admin still sees the host's paths");
+  });
+});

server/routes/settings.js

@@ -35,7 +35,7 @@ const {
   importExportBundle,
   exportableRun,
 } = require("../lib/data-transfer");
-const { sessionWhere, bySessionWhere } = require("../lib/scope");
+const { sessionWhere, bySessionWhere, memberId } = require("../lib/scope");
 
 function getDbSize() {
   try {
@@ -127,15 +127,20 @@ router.get("/info", (req, res) => {
     h1: getCount(60 * 60 * 1000),
   };
 
+  // A developer (server mode) gets the numbers, never a path of the host's:
+  // the transcript cache is keyed by EVERY user's transcript path (their
+  // folders, their session ids), and the DB/settings paths are the host's.
+  const developer = memberId(req) != null;
+  const cache = transcriptCache.stats();
   res.json({
     db: {
-      path: DB_PATH,
+      path: developer ? null : DB_PATH,
       size: dbSize,
       counts,
       pragmas,
       load_stats,
     },
-    hooks: hookStatus,
+    hooks: developer ? { installed: hookStatus.installed } : hookStatus,
     server: {
       version: APP_VERSION,
       uptime: process.uptime(),
@@ -152,7 +157,7 @@ router.get("/info", (req, res) => {
       free_mem: os.freemem(),
       cpus: os.cpus().length,
     },
-    transcript_cache: transcriptCache.stats(),
+    transcript_cache: developer ? { ...cache, keys: undefined } : cache,
   });
 });
 

server/routes/updates.js

@@ -24,9 +24,11 @@ const { schedulePluginUpdate } = require("../lib/plugin-update");
 
 const router = Router();
 
-router.get("/status", async (_req, res) => {
+router.get("/status", async (req, res) => {
   try {
     const status = await getUpdatesStatus();
+    // The install folder is the host's; a developer sees versions only.
+    if (require("../lib/scope").memberId(req) != null) status.repo_root = null;
     res.json(status);
   } catch (err) {
     res.status(500).json({