wealthbox-cli 1.1.4__tar.gz → 1.1.6__tar.gz

{wealthbox_cli-1.1.4 → wealthbox_cli-1.1.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: wealthbox-cli
-Version: 1.1.4
+Version: 1.1.6
 Summary: CLI and client library for the Wealthbox CRM API
 Project-URL: Homepage, https://github.com/massive-value/wealthbox-cli
 Project-URL: Documentation, https://massive-value.github.io/wealthbox-cli/
@@ -144,11 +144,13 @@ wbox config set-token        # paste your Wealthbox API token
 wbox skills install          # interactive: pick Claude Code, Codex, or both
 ```
 
-The installer asks which platforms to target, copies the skill into the right directory (`~/.claude/skills/` for Claude Code, `~/.codex/skills/` for Codex — with `SKILL.md` renamed to `AGENTS.md` automatically), and offers to bootstrap your firm's customizations from the Wealthbox API.
+The installer asks which platforms to target, copies the skill into the right directory (`~/.claude/skills/` for Claude Code, `~/.codex/skills/` for Codex — both use `SKILL.md`), and offers to bootstrap your firm's customizations from the Wealthbox API.
 
-### First agent run bootstraps itself
+> **Upgrading from < 1.1.6?** Earlier versions installed the codex skill as `AGENTS.md`. That was wrong — codex uses `SKILL.md` for skills (and `AGENTS.md` only as the project-level instructions file, in place of `CLAUDE.md`). If your codex install still has `AGENTS.md`, reinstall with `wbox skills install --platform codex --force`.
 
-The first time an agent invokes the skill, it walks you through populating firm-specific conventions — category values, custom fields, users, defaults, required fields, household naming, named workflows. No further CLI commands needed. After that first run, the bootstrap artifacts self-delete from the skill.
+### First agent run captures firm conventions
+
+`wbox skills bootstrap` populates the API-derived files (categories, custom fields, users) — that's the *quantitative* half. The *qualitative* half (firm defaults, naming conventions, named workflows, required fields) happens the first time an agent invokes the skill: it walks you through the questions in `bootstrap.md` and stamps `firm.onboarded_at` in `_meta.json` when done. On subsequent runs the bootstrap path is skipped automatically.
 
 ### Refresh after firm changes
 
@@ -167,6 +169,7 @@ wbox skills list             # show where the skill is installed + last bootstra
 wbox skills doctor           # diagnose install state + token
 wbox skills sync             # copy firm/ from one install to others (Claude Code <-> Codex)
 wbox skills upgrade          # refresh template files (SKILL.md, references/, ...) preserving firm/
+wbox skills mark-onboarded   # stamp firm.onboarded_at after qualitative Q&A (the agent calls this for you)
 wbox skills uninstall        # remove the skill
 ```
 

{wealthbox_cli-1.1.4 → wealthbox_cli-1.1.6}/README.md

@@ -103,11 +103,13 @@ wbox config set-token        # paste your Wealthbox API token
 wbox skills install          # interactive: pick Claude Code, Codex, or both
 ```
 
-The installer asks which platforms to target, copies the skill into the right directory (`~/.claude/skills/` for Claude Code, `~/.codex/skills/` for Codex — with `SKILL.md` renamed to `AGENTS.md` automatically), and offers to bootstrap your firm's customizations from the Wealthbox API.
+The installer asks which platforms to target, copies the skill into the right directory (`~/.claude/skills/` for Claude Code, `~/.codex/skills/` for Codex — both use `SKILL.md`), and offers to bootstrap your firm's customizations from the Wealthbox API.
 
-### First agent run bootstraps itself
+> **Upgrading from < 1.1.6?** Earlier versions installed the codex skill as `AGENTS.md`. That was wrong — codex uses `SKILL.md` for skills (and `AGENTS.md` only as the project-level instructions file, in place of `CLAUDE.md`). If your codex install still has `AGENTS.md`, reinstall with `wbox skills install --platform codex --force`.
 
-The first time an agent invokes the skill, it walks you through populating firm-specific conventions — category values, custom fields, users, defaults, required fields, household naming, named workflows. No further CLI commands needed. After that first run, the bootstrap artifacts self-delete from the skill.
+### First agent run captures firm conventions
+
+`wbox skills bootstrap` populates the API-derived files (categories, custom fields, users) — that's the *quantitative* half. The *qualitative* half (firm defaults, naming conventions, named workflows, required fields) happens the first time an agent invokes the skill: it walks you through the questions in `bootstrap.md` and stamps `firm.onboarded_at` in `_meta.json` when done. On subsequent runs the bootstrap path is skipped automatically.
 
 ### Refresh after firm changes
 
@@ -126,6 +128,7 @@ wbox skills list             # show where the skill is installed + last bootstra
 wbox skills doctor           # diagnose install state + token
 wbox skills sync             # copy firm/ from one install to others (Claude Code <-> Codex)
 wbox skills upgrade          # refresh template files (SKILL.md, references/, ...) preserving firm/
+wbox skills mark-onboarded   # stamp firm.onboarded_at after qualitative Q&A (the agent calls this for you)
 wbox skills uninstall        # remove the skill
 ```
 

{wealthbox_cli-1.1.4 → wealthbox_cli-1.1.6}/docs/changelog.md

@@ -6,6 +6,30 @@ This project uses [Semantic Versioning](https://semver.org/).
 
 ---
 
+## [1.1.6](https://github.com/massive-value/wealthbox-cli/releases/tag/v1.1.6) — 2026-05-01
+
+### Fixed
+- `wbox skills bootstrap` and `wbox skills refresh` now paginate every API call. Previously they fetched a single page (default `per_page=25`) for each category type, custom-field document type, and the user list, silently truncating any workspace with more than 25 entries per group. Generated `firm/categories.md`, `firm/custom-fields.md`, and `firm/users.md` are now complete.
+- The codex skill installs as `SKILL.md`, not `AGENTS.md`. Codex uses `SKILL.md` for skills (same as Claude Code); `AGENTS.md` is the project-level instructions file (the codex equivalent of `CLAUDE.md`) and was never the right name for a skill. **Migration:** if you installed the skill under codex on 1.1.5 or earlier, reinstall with `wbox skills install --platform codex --force` to drop the stale `AGENTS.md`.
+- The skill's First Run check now keys on `firm.onboarded_at` instead of the mere presence of a `firm` section. Previously, running `wbox skills bootstrap` populated the `firm` section, which made the agent skip the qualitative Q&A in `bootstrap.md` on first invocation. The CLI bootstrap is the *quantitative* half (categories, custom fields, users); the *qualitative* half (firm defaults, naming conventions, named workflows) happens at first agent invocation, and the agent now stamps `firm.onboarded_at` when it finishes.
+
+### Added
+- `wbox skills mark-onboarded` — stamps `firm.onboarded_at` in the skill's `_meta.json`. The agent invokes this as the last step of `bootstrap.md`; users typically don't need to run it manually.
+- `wbox skills list` and `wbox skills doctor` now report onboarded status (`bootstrapped (qualitative pending)` vs `onboarded`).
+
+---
+
+## [1.1.5](https://github.com/massive-value/wealthbox-cli/releases/tag/v1.1.5) — 2026-05-01
+
+### Fixed
+- `wbox skills install`, `wbox skills bootstrap`, and `wbox skills refresh` now honor every token source the rest of the CLI does. Previously these commands only looked at the `--token` flag and `WEALTHBOX_TOKEN` env var, ignoring tokens stored via `wbox config set-token` and `.env` files. Users running `wbox skills install` after `wbox config set-token` saw "Wealthbox token required" errors despite having a token set.
+- The "token required" error message now lists every supported source (flag, env var, config file, `.env`).
+
+### Added
+- `wbox skills install` accepts `--token` (matching `bootstrap`, `refresh`, and `doctor`), so a token can be supplied inline for the post-install bootstrap step.
+
+---
+
 ## [1.1.4](https://github.com/massive-value/wealthbox-cli/releases/tag/v1.1.4) — 2026-05-01
 
 ### Added

{wealthbox_cli-1.1.4 → wealthbox_cli-1.1.6}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "wealthbox-cli"
-version = "1.1.4"
+version = "1.1.6"
 description = "CLI and client library for the Wealthbox CRM API"
 readme = "README.md"
 requires-python = ">=3.11"

{wealthbox_cli-1.1.4 → wealthbox_cli-1.1.6}/src/wealthbox_tools/cli/_skill_bootstrap.py

@@ -176,17 +176,41 @@ def update_firm_meta(
     cli_version: str,
     generated_files: list[str],
 ) -> None:
-    """Set the firm section of _meta.json. Preserves any existing template section."""
+    """Set the firm section of _meta.json. Preserves the qualitative onboarded_at marker
+    (set by the agent via `wbox skills mark-onboarded`) and any existing template section."""
     now = datetime.now(timezone.utc).isoformat()
     meta = read_meta(skill_dir)
-    meta["firm"] = {
+    existing_firm = meta.get("firm") or {}
+    firm_section: dict[str, Any] = {
         "identity": identity,
         "cli_version": cli_version,
         "files": {name: now for name in generated_files},
     }
+    onboarded_at = existing_firm.get("onboarded_at")
+    if onboarded_at:
+        firm_section["onboarded_at"] = onboarded_at
+    meta["firm"] = firm_section
     write_meta(skill_dir, meta)
 
 
+def mark_firm_onboarded(skill_dir: Path) -> str:
+    """Stamp `firm.onboarded_at` with the current UTC time. Returns the timestamp written.
+
+    Raises FileNotFoundError if no `firm` section exists yet (caller must run
+    `wbox skills bootstrap` before marking onboarded)."""
+    meta = read_meta(skill_dir)
+    firm = meta.get("firm")
+    if not firm:
+        raise FileNotFoundError(
+            f"{meta_path(skill_dir)} has no 'firm' section; run 'wbox skills bootstrap' first."
+        )
+    now = datetime.now(timezone.utc).isoformat()
+    firm["onboarded_at"] = now
+    meta["firm"] = firm
+    write_meta(skill_dir, meta)
+    return now
+
+
 def copy_firm_meta(src_skill_dir: Path, dst_skill_dir: Path) -> bool:
     """Copy the firm section of _meta.json from src to dst, preserving dst's template.
 
@@ -212,6 +236,8 @@ from importlib.metadata import version as _pkg_version  # noqa: E402
 from wealthbox_tools.client import WealthboxClient  # noqa: E402
 from wealthbox_tools.models import CategoryListQuery, CategoryType, DocumentType  # noqa: E402
 
+from ._util import get_client  # noqa: E402
+
 
 @dataclass
 class BootstrapResult:
@@ -228,18 +254,18 @@ async def _fetch_all(
     for ct in CategoryType:
         if ct is CategoryType.CUSTOM_FIELDS:
             continue
-        resp = await client.list_categories(ct)
+        resp = await client.list_all_categories(ct)
         categories[ct.value] = resp.get(ct.value, [])
 
     custom_fields: dict[str, list[dict[str, Any]]] = {}
     for dt in DocumentType:
-        resp = await client.list_categories(
+        resp = await client.list_all_categories(
             CategoryType.CUSTOM_FIELDS,
             CategoryListQuery(document_type=dt),
         )
         custom_fields[dt.value] = resp.get("custom_fields", [])
 
-    users_resp = await client.list_users()
+    users_resp = await client.list_all_users()
     users = users_resp.get("users", [])
 
     me = await client.get_me()
@@ -264,7 +290,7 @@ def bootstrap_skill_dir(
     firm_dir.mkdir(parents=True, exist_ok=True)
 
     async def _run() -> tuple[dict, dict, list, dict]:
-        async with WealthboxClient(token=token) as client:
+        async with get_client(token) as client:
             return await _fetch_all(client)
 
     categories, custom_fields, users, firm_identity = asyncio.run(_run())

{wealthbox_cli-1.1.4 → wealthbox_cli-1.1.6}/src/wealthbox_tools/cli/_skill_platforms.py

@@ -40,7 +40,7 @@ def detect_platforms() -> list[Platform]:
             id="codex",
             label="Codex",
             root_dir=home / ".codex" / "skills",
-            skill_filename="AGENTS.md",
+            skill_filename="SKILL.md",
             requires_project_cwd=False,
         ),
     ]
@@ -69,7 +69,6 @@ def install_skill(platform: Platform, src: Path, *, force: bool) -> None:
         _safe_rmtree(platform, dest)
     dest.parent.mkdir(parents=True, exist_ok=True)
     shutil.copytree(src, dest)
-    _rename_skill_md(platform, dest)
 
 
 def upgrade_skill(platform: Platform, src: Path) -> None:
@@ -85,20 +84,6 @@ def upgrade_skill(platform: Platform, src: Path) -> None:
             f"Cannot upgrade: nothing installed at {dest}"
         )
     shutil.copytree(src, dest, dirs_exist_ok=True)
-    _rename_skill_md(platform, dest)
-
-
-def _rename_skill_md(platform: Platform, dest: Path) -> None:
-    """Rename SKILL.md → platform.skill_filename if needed (e.g. AGENTS.md for codex)."""
-    if platform.skill_filename == "SKILL.md":
-        return
-    skill_md = dest / "SKILL.md"
-    target = dest / platform.skill_filename
-    if not skill_md.exists():
-        return
-    if target.exists():
-        target.unlink()
-    skill_md.rename(target)
 
 
 def uninstall_skill(platform: Platform) -> None:

{wealthbox_cli-1.1.4 → wealthbox_cli-1.1.6}/src/wealthbox_tools/cli/_util.py

@@ -62,6 +62,12 @@ def get_client(token: str | None = None) -> WealthboxClient:
             token = os.environ.get("WEALTHBOX_TOKEN")
         except ImportError:
             pass
+    if not token:
+        raise ValueError(
+            "Wealthbox token required. Provide one via any of: "
+            "--token flag, WEALTHBOX_TOKEN env var, "
+            "`wbox config set-token`, or a .env file in the working directory."
+        )
     return WealthboxClient(token=token)
 
 

{wealthbox_cli-1.1.4 → wealthbox_cli-1.1.6}/src/wealthbox_tools/cli/skills.py

@@ -8,7 +8,7 @@ from pathlib import Path
 
 import typer
 
-from ._skill_bootstrap import copy_firm_meta, read_meta, update_template_meta
+from ._skill_bootstrap import copy_firm_meta, mark_firm_onboarded, read_meta, update_template_meta
 from ._skill_platforms import (
     Platform,
     SkillInstallError,
@@ -29,13 +29,14 @@ app = typer.Typer(
 
 @app.command("list", help="Show where the skill is installed per platform.")
 def list_platforms() -> None:
-    header = ("platform", "path", "status", "template-version", "last-bootstrap")
-    rows: list[tuple[str, str, str, str, str]] = []
+    header = ("platform", "path", "status", "template-version", "last-bootstrap", "onboarded")
+    rows: list[tuple[str, str, str, str, str, str]] = []
     for p in detect_platforms():
         dest = skill_dir(p)
         installed = is_installed(p)
         template_version = "-"
         last_bootstrap = "-"
+        onboarded = "-"
         if installed:
             meta = read_meta(dest)
             template_version = meta.get("template", {}).get("cli_version", "-")
@@ -43,12 +44,14 @@ def list_platforms() -> None:
             firm_files = firm_section.get("files") or {}
             if firm_files:
                 last_bootstrap = max(firm_files.values())
+            onboarded = firm_section.get("onboarded_at") or "no"
         rows.append((
             p.id,
             str(dest),
             "installed" if installed else "not installed",
             template_version,
             last_bootstrap,
+            onboarded,
         ))
 
     widths = [
@@ -98,6 +101,7 @@ def install_cmd(
     ),
     force: bool = typer.Option(False, "--force", help="Overwrite existing install."),
     no_bootstrap: bool = typer.Option(False, "--no-bootstrap", help="Skip the post-install bootstrap prompt."),
+    token: str | None = typer.Option(None, envvar="WEALTHBOX_TOKEN", hidden=True),
 ) -> None:
     targets = _resolve_platforms(platforms_flag) if platforms_flag else _prompt_platforms()
 
@@ -122,7 +126,7 @@ def install_cmd(
         if typer.confirm("Run 'wbox skills bootstrap' now?", default=True):
             from ._skill_bootstrap import bootstrap_skill_dir
             for target in targets:
-                bootstrap_skill_dir(skill_dir(target), token=None, generated_only=False)
+                bootstrap_skill_dir(skill_dir(target), token=token, generated_only=False)
                 typer.echo(f"OK bootstrapped {target.id}")
 
 
@@ -267,13 +271,19 @@ def doctor_cmd(
     for p in detect_platforms():
         status = "installed" if is_installed(p) else "not installed"
         meta = read_meta(skill_dir(p)) if is_installed(p) else {}
-        bootstrapped = "bootstrapped" if meta.get("firm") else "not bootstrapped"
+        firm_section = meta.get("firm") or {}
+        if not firm_section:
+            firm_state = "not bootstrapped"
+        elif firm_section.get("onboarded_at"):
+            firm_state = "onboarded"
+        else:
+            firm_state = "bootstrapped (qualitative pending)"
         template_v = (meta.get("template") or {}).get("cli_version", "-")
         upgrade_hint = ""
         if template_v not in ("-", current):
             upgrade_hint = f"  [upgrade available: {template_v} -> {current}]"
         typer.echo(
-            f"  {p.id:<22} {status:<16} {bootstrapped:<18} "
+            f"  {p.id:<22} {status:<16} {firm_state:<36} "
             f"template={template_v:<10} {skill_dir(p)}{upgrade_hint}"
         )
 
@@ -441,6 +451,41 @@ def sync_cmd(
         )
 
 
+@app.command(
+    "mark-onboarded",
+    help="Stamp firm.onboarded_at in _meta.json. Run after qualitative firm Q&A is captured.",
+)
+def mark_onboarded_cmd(
+    platforms_flag: list[str] = typer.Option(
+        [], "--platform", "-p",
+        help="Platform id. Default: every installed platform.",
+    ),
+) -> None:
+    if platforms_flag:
+        targets = _resolve_platforms(platforms_flag)
+        for t in targets:
+            if not is_installed(t):
+                typer.echo(f"Error: {t.id!r} is not installed.", err=True)
+                raise typer.Exit(code=2)
+    else:
+        targets = [p for p in detect_platforms() if is_installed(p)]
+
+    if not targets:
+        typer.echo(
+            "No installed platforms found. Run 'wbox skills install' first.",
+            err=True,
+        )
+        raise typer.Exit(code=2)
+
+    for t in targets:
+        try:
+            ts = mark_firm_onboarded(skill_dir(t))
+        except FileNotFoundError as e:
+            typer.echo(f"Error: {t.id}: {e}", err=True)
+            raise typer.Exit(code=2) from e
+        typer.echo(f"OK marked {t.id} onboarded at {ts}")
+
+
 @app.command("uninstall", help="Remove the wealthbox-crm skill from a platform.")
 def uninstall_cmd(
     platforms_flag: list[str] = typer.Option([], "--platform", "-p"),

wealthbox_cli-1.1.6/src/wealthbox_tools/client/categories.py

@@ -0,0 +1,30 @@
+from __future__ import annotations
+
+from collections.abc import Callable
+from typing import Any
+
+from wealthbox_tools.models import CategoryListQuery, CategoryType
+
+
+class CategoriesMixin:
+    """Categories Resources"""
+
+    async def list_categories(self, category: CategoryType, query: CategoryListQuery | None = None) -> dict[str, Any]:
+        params = query.model_dump(exclude_none=True) if query else None
+        resp = await self._request("GET", f"/categories/{category}", params=params)  # type: ignore[attr-defined]
+        return resp.json()
+
+    async def list_all_categories(
+        self,
+        category: CategoryType,
+        query: CategoryListQuery | None = None,
+        on_progress: Callable[[int, int], None] | None = None,
+    ) -> dict[str, Any]:
+        """Fetch every page of a category type. Custom fields use collection key 'custom_fields'."""
+        params = query.model_dump(exclude_none=True) if query else {}
+        params.pop("page", None)
+        params.pop("per_page", None)
+        collection_key = "custom_fields" if category is CategoryType.CUSTOM_FIELDS else category.value
+        return await self.fetch_all_pages(  # type: ignore[attr-defined]
+            f"/categories/{category}", params, collection_key, on_progress=on_progress
+        )

wealthbox_cli-1.1.6/src/wealthbox_tools/client/users.py

@@ -0,0 +1,28 @@
+from __future__ import annotations
+
+from collections.abc import Callable
+from typing import Any
+
+from wealthbox_tools.models.common import PaginationQuery
+
+
+class UsersMixin:
+    """Users Resource"""
+
+    async def list_users(self, query: PaginationQuery | None = None) -> dict[str, Any]:
+        params = query.model_dump(exclude_none=True) if query else None
+        resp = await self._request("GET", "/users", params=params)  # type: ignore[attr-defined]
+        return resp.json()
+
+    async def list_all_users(
+        self,
+        query: PaginationQuery | None = None,
+        on_progress: Callable[[int, int], None] | None = None,
+    ) -> dict[str, Any]:
+        """Fetch every page of users. Use when full enumeration is required (e.g. firm bootstrap)."""
+        params = query.model_dump(exclude_none=True) if query else {}
+        params.pop("page", None)
+        params.pop("per_page", None)
+        return await self.fetch_all_pages(  # type: ignore[attr-defined]
+            "/users", params, "users", on_progress=on_progress
+        )

{wealthbox_cli-1.1.4 → wealthbox_cli-1.1.6}/src/wealthbox_tools/skills/wealthbox-crm/SKILL.md

@@ -10,14 +10,20 @@ opportunities, projects, workflows, households, and read-only lookups (categorie
 
 ## First Run
 
-If `_meta.json` in this skill's directory does **not** contain a `firm`
-section, the firm has not been onboarded yet on this machine. Before
-handling the user's request:
+If `_meta.json.firm.onboarded_at` is missing in this skill's directory,
+the firm has not been onboarded yet on this machine. Before handling the
+user's request:
 
 1. Read `bootstrap.md` in this skill's directory.
 2. Follow it end-to-end.
-3. The bootstrap will populate `_meta.json.firm` so this check passes
-   on subsequent runs.
+3. The final step writes `_meta.json.firm.onboarded_at`, which makes
+   this check pass on subsequent runs.
+
+`firm.onboarded_at` is the qualitative-onboarding marker. The CLI
+bootstrap (`wbox skills bootstrap`) only populates the API-derived
+parts of `firm` (identity, generated files); it intentionally does
+*not* set `onboarded_at`, because that requires the per-firm Q&A in
+`bootstrap.md`.
 
 ## Staleness Check
 

{wealthbox_cli-1.1.4 → wealthbox_cli-1.1.6}/src/wealthbox_tools/skills/wealthbox-crm/bootstrap.md

@@ -1,11 +1,14 @@
 # First-Run Bootstrap
 
-You were told to read this because `_meta.json` in the skill root does not
-contain a `firm` section. Follow every step below.
+You were told to read this because `_meta.json.firm.onboarded_at` is missing
+in the skill root. Follow every step below, then run the marker command in
+Step 5 so this bootstrap path is skipped on future invocations.
 
 ## Step 1 — Fetch API-derived firm data
 
-Run this command:
+Check `_meta.json` first. If it already has a `firm` section with `identity`
+and a `files` map, the CLI bootstrap has already run on this machine and
+you can skip to Step 2. Otherwise, run:
 
 ```bash
 wbox skills bootstrap
@@ -13,11 +16,12 @@ wbox skills bootstrap
 
 It populates these files per installed platform:
 
-- `firm/categories.md` — every category type + valid values
-- `firm/custom-fields.md` — custom fields per document type + valid values
-- `firm/users.md` — user id → name → email
+- `firm/categories.md` — every category type + valid values (paginated; complete)
+- `firm/custom-fields.md` — custom fields per document type + valid values (paginated; complete)
+- `firm/users.md` — user id → name → email (paginated; complete)
 - `_meta.json` (skill root) — adds the `firm` section: identity, refresh
-  timestamps, and the CLI version that ran the bootstrap
+  timestamps, and the CLI version that ran the bootstrap. Does **not**
+  set `onboarded_at` — that's Step 5.
 - Stub files: `firm/contacts.md`, `firm/tasks.md`, `firm/notes.md`,
   `firm/events.md`, `firm/opportunities.md`, `firm/projects.md`,
   `firm/workflows.md` — one-line TODOs you will fill in next.
@@ -32,6 +36,10 @@ in `firm/<resource>.md`, replacing the stub TODO.
 
 Read `firm/categories.md` and `firm/custom-fields.md` first so every
 multiple-choice prompt offers **real** values from the firm's workspace.
+If the advisor mentions a tag, custom-field option, or user that you can't
+find in those files, flag it — they're generated from a paginated fetch and
+*should* be complete. A missing entry usually means the value doesn't exist
+in the workspace yet (rather than a pagination gap).
 
 ### firm/contacts.md
 
@@ -90,11 +98,22 @@ Ask the advisor: "Is there any firm convention, custom process, or preference
 I haven't asked about that I should follow?" Append freeform answers to
 `firm/workflows.md` under a `## Other Conventions` heading.
 
-## Step 4 — Confirm and continue
+## Step 4 — Mark the firm onboarded
+
+Run:
+
+```bash
+wbox skills mark-onboarded
+```
+
+This stamps `firm.onboarded_at` in `_meta.json`. The First Run check in
+`SKILL.md` keys off this field, so future invocations will skip the
+bootstrap path automatically.
+
+## Step 5 — Confirm and continue
 
 Tell the advisor: "First-run setup complete — `firm/` populated for <firm
 name>." Then return to the user's original request.
 
 You don't need to delete this file or trim `SKILL.md`. The first-run check
-keys on `_meta.json.firm`, which now exists, so the bootstrap path will be
-skipped on every subsequent invocation automatically.
+keys on `_meta.json.firm.onboarded_at`, which now exists.

wealthbox_cli-1.1.4/src/wealthbox_tools/client/categories.py

@@ -1,14 +0,0 @@
-from __future__ import annotations
-
-from typing import Any
-
-from wealthbox_tools.models import CategoryListQuery, CategoryType
-
-
-class CategoriesMixin:
-    """Categories Resources"""
-
-    async def list_categories(self, category: CategoryType, query: CategoryListQuery | None = None) -> dict[str, Any]:
-        params = query.model_dump(exclude_none=True) if query else None
-        resp = await self._request("GET", f"/categories/{category}", params=params)  # type: ignore[attr-defined]
-        return resp.json()

wealthbox_cli-1.1.4/src/wealthbox_tools/client/users.py

@@ -1,14 +0,0 @@
-from __future__ import annotations
-
-from typing import Any
-
-from wealthbox_tools.models.common import PaginationQuery
-
-
-class UsersMixin:
-    """Users Resource"""
-    
-    async def list_users(self, query: PaginationQuery | None = None) -> dict[str, Any]:
-        params = query.model_dump(exclude_none=True) if query else None
-        resp = await self._request("GET", "/users", params=params)  # type: ignore[attr-defined]
-        return resp.json()