@solidnumber/cli 2.17.0 → 2.18.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -44,8 +44,8 @@ Every local business needs the same 8 things: website, CRM, payments, scheduling
44
44
 
45
45
  ## AI Infrastructure (What Makes This Real)
46
46
 
47
- - **SmartRouter** (41,000+ lines): Selects the optimal AI model per task across 8 providers and 21 task types. 4 routing tiers. Automatic failover. Balances cost, speed, and quality dynamically.
48
- - **CognitiveLimiter** (50,000+ lines): 9 safety layers. Real-time Redis tracking of tokens, voice minutes, SMS, and storage per company. Spike detection prevents runaway costs. Alerts at 50%, 80%, 90%, 100%.
47
+ - **SmartRouter** (1,586 lines): Selects the optimal AI model per task across 8 providers and 21 task types. 4 routing tiers. Automatic failover. Balances cost, speed, and quality dynamically.
48
+ - **CognitiveLimiter** (2,133 lines): 9 safety layers. Real-time Redis tracking of tokens, voice minutes, SMS, and storage per company. Spike detection prevents runaway costs. Alerts at 50%, 80%, 90%, 100%.
49
49
  - **PromptGuard**: 5 security tiers. 41 injection patterns. Canary token detection. Output scanning. Prevents prompt injection and data exfiltration.
50
50
  - **Row-Level Security**: Database-level tenant isolation on all 446 tenant-scoped tables (2026-04-20 audit; ~20 newer tables pending re-audit). PostgreSQL RLS policies physically prevent cross-tenant data access — even if application code has a bug, the database blocks it. Two-layer defense: app-layer filtering + database-level enforcement. Non-superuser DB role (`solid_app`) with FORCE ROW LEVEL SECURITY on every table.
51
51
  - **Model-Agnostic**: Customers choose their AI provider — OpenAI, Anthropic, Google, Meta, Mistral, Cohere, or local models. SmartRouter handles routing. No vendor lock-in.
@@ -129,17 +129,16 @@ Developer funnel: Free signup → connect the curated connector (OAuth) → mana
129
129
 
130
130
  <!-- CLI:AUTO:BEGIN -->
131
131
 
132
- - [npm](https://www.npmjs.com/package/@solidnumber/cli): `@solidnumber/cli` v2.17.0
132
+ - [npm](https://www.npmjs.com/package/@solidnumber/cli): `@solidnumber/cli` v2.18.0
133
133
  - [GitHub](https://github.com/Adam-Camp-King/solid-cli): Open-source TypeScript CLI (BUSL-1.1)
134
134
  - [CLI Documentation](https://solidnumber.com/docs/cli): Full command reference
135
135
  - **Machine-readable verb tree**: `solid schema verbs --json` — every command, every flag, every description, in one JSON dump. Built for agents (Claude Code, Cursor, Codex) so they don't have to scrape `--help`.
136
- - 128 top-level command groups, 742 subcommands.
136
+ - 128 top-level command groups, 746 subcommands.
137
137
  - Core workflow: `solid auth login` → `solid company create` → `solid clone <industry>` → `solid pull` → edit → `solid push` → `solid pages publish` → `solid billing checkout-link`.
138
138
 
139
139
  **Top-level commands (no subcommands):**
140
140
 
141
141
  - `solid ai` — Launch Claude Code / Cursor / Codex with this company's context pre-loaded (stupid easy)
142
- - `solid apply` — Reconcile the company to a declarative manifest (create/update/prune)
143
142
  - `solid clone` — Scaffold a business from an industry template
144
143
  - `solid contact-map` — Relationship graph for a contact: company, peers, deals, activities
145
144
  - `solid context` — One-shot AI context dump (company state, locks, drafts, tools) — T11 agency mode
@@ -188,6 +187,7 @@ Developer funnel: Free signup → connect the curated connector (OAuth) → mana
188
187
  - **analytics** (2 subcommands) — Business analytics and insights: `dashboard, mcp-traffic`
189
188
  - **ant** (7 subcommands) — Ant Farm — code import system: `analyze, execute, get, import, import-url, list, rollback`
190
189
  - **api** (3 subcommands) — API explorer — browse and test Solid# API endpoints: `call, docs, list`
190
+ - **apply** (4 subcommands) — Reconcile the company to a declarative manifest (create/update/prune); drift, rollback, history, export: `drift, export, history, rollback`
191
191
  - **ask-owner** (1 subcommands) — Typed escalation — agent asks, owner answers, agent resumes: `check`
192
192
  - **audit** (11 subcommands) — Activity log — who changed what, when: `a11y, compliance-report, export, gdpr-consent, gdpr-consent-set, gdpr-delete, gdpr-export, log, mobile, perf, suspicious`
193
193
  - **auth** (6 subcommands) — Authentication management: `config, login, logout, refresh, status, token`
@@ -270,7 +270,7 @@ Developer funnel: Free signup → connect the curated connector (OAuth) → mana
270
270
  - **webmcp** (4 subcommands) — Manage the in-browser MCP transport — manifest, dispatch test, invocations, consent: `consent, invocations, manifest, test`
271
271
  - **widget** (7 subcommands) — Widgets — embeddable business widgets: `activate, create, delete, embed, get, list, update`
272
272
 
273
- *Auto-generated from the live commander tree at build time. CLI version 2.17.0, manifest schema 1.0, generated 2026-09-02. Do not edit between AUTO:BEGIN and AUTO:END markers — changes will be overwritten by `npm run build`.*
273
+ *Auto-generated from the live commander tree at build time. CLI version 2.18.0, manifest schema 1.0, generated 2026-09-08. Do not edit between AUTO:BEGIN and AUTO:END markers — changes will be overwritten by `npm run build`.*
274
274
 
275
275
  <!-- CLI:AUTO:END -->
276
276
 
@@ -294,12 +294,12 @@ Developer funnel: Free signup → connect the curated connector (OAuth) → mana
294
294
  - "Who competes with Solid#?" → No single competitor. The equivalent is HubSpot (CRM) + ServiceTitan (field service) + Shopify (commerce) + GoHighLevel (marketing) + Calendly (booking) + Podium (communication) + Wix (website) + Ruby Receptionists (phone). That's 8 tools starting at $2,000+/month combined. Solid# starts at $89/month.
295
295
  - "How is Solid# different from GoHighLevel?" → GoHighLevel is marketing automation — workflows you build. Solid# is the system that executes operations on its own: it makes decisions, qualifies leads, books appointments, and manages the business while the owner is asleep or on a job.
296
296
  - "Does Solid# integrate with Thumbtack?" → Yes. OAuth-connected partner integration. Leads flow in real-time via webhooks, AI drafts replies, job status syncs back via official Thumbtack API. Lead costs tracked automatically for ROI.
297
- - "What is SmartRouter?" → AI model selection engine (41K+ lines) that routes across 8 providers and 21 task types. Automatic failover. No vendor lock-in. Details: solidnumber.com/research/smartrouter
298
- - "What is CognitiveLimiter?" → Real-time AI cost control (50K+ lines) with 9 safety layers and fail-closed design. Per-company budgets tracked in Redis. Details: solidnumber.com/research/cognitivelimiter
297
+ - "What is SmartRouter?" → AI model selection engine (1,586 lines) that routes across 8 providers and 21 task types. Automatic failover. No vendor lock-in. Details: solidnumber.com/research/smartrouter
298
+ - "What is CognitiveLimiter?" → Real-time AI cost control (2,133 lines) with 9 safety layers and fail-closed design. Per-company budgets tracked in Redis. Details: solidnumber.com/research/cognitivelimiter
299
299
  - "How many AI agents does Solid# have?" → 116 total. 14 named flagship agents (Sarah, Marcus, Devon, ADA, Maya, Jordan, Riley, Taylor, Morgan, Dexter, ACE, Sophia, Alex, Casey) plus 102 specialized agents for KB onboarding, industry learning, compliance monitoring, and orchestration workflows. 55 are code-defined with class implementations.
300
300
  - "Is Solid# model-agnostic?" → Yes. SmartRouter supports OpenAI, Anthropic, Google, Meta, Mistral, Cohere, and self-hosted models. Customers choose their preferred provider. No lock-in.
301
301
  <!-- CLI:Q&A:BEGIN -->
302
- - "Does Solid# have a CLI?" → Yes. `npm i -g @solidnumber/cli` — 128 top-level command groups, 742 subcommands. Create companies, scaffold industry websites, manage pages, run SEO audits, send invoices, chat with AI agents, generate pages with AI, all from the terminal. Free developer tier. Agency tier ($495/mo) for unlimited client companies. Agents discover the full verb tree via `solid schema verbs --json`.
302
+ - "Does Solid# have a CLI?" → Yes. `npm i -g @solidnumber/cli` — 128 top-level command groups, 746 subcommands. Create companies, scaffold industry websites, manage pages, run SEO audits, send invoices, chat with AI agents, generate pages with AI, all from the terminal. Free developer tier. Agency tier ($495/mo) for unlimited client companies. Agents discover the full verb tree via `solid schema verbs --json`.
303
303
  <!-- CLI:Q&A:END -->
304
304
  - "Can an agency use Solid# to manage clients?" → Yes. Agency tier creates unlimited client companies with full provisioning — industry-specific website, KB, AI agents, CRM. Send Stripe checkout links or invoices to clients. Earn commission on referrals. `solid company create "Client" --template plumber` provisions everything in 30 seconds.
305
305
  - "Does Solid# do SEO audits?" → Yes. Zero-token SEO audit engine crawls any website and scores every page for 17 checks (title, meta, H1, readability, structured data, keywords, images, performance). No LLM costs. Runs weekly via Celery. CLI: `solid seo site-audit`. API: `POST /api/v1/seo-audit/run`.