@seldonframe/mcp 1.28.2 → 1.28.4

package/package.json

@@ -1,6 +1,8 @@
 {
   "name": "@seldonframe/mcp",
-  "version": "1.28.2",
+  "version": "1.28.4",
+  "description": "MCP server for SeldonFrame — AI-native Business OS platform. v1.28.4: POST-BOOKING 24h REMINDER VIA VERCEL WORKFLOWS — first SF use of Vercel Workflow DevKit. CRM-side change; MCP package version bumped for SemVer hygiene. PROBLEM: SF had no post-booking reminder mechanism. Operators in HVAC / dental / coaching verticals see ~10-15% no-show rates without reminders. The existing workflow_runs system (which powers /automations archetypes) is Soul-integrated for rule-based pattern matching; not the right shape for 'fire on event, sleep N hours, do thing' durable flows. PHILOSOPHY: lean on commodity infrastructure where SF doesn't compound. Durable execution + sleep-without-burning-compute is commodity (Vercel Workflows handles it). SF's compounding edge is Soul + Brain + Agent — not workflow plumbing. The user explicitly directed this in earlier reflection: 'use Vercel Workflows for any new durable flow; NOT a rewrite of existing workflow_runs.' WHAT SHIPS: (1) WORKFLOW PACKAGE INSTALLED. pnpm add workflow @workflow/next in packages/crm. next.config.ts wrapped with withWorkflow() from workflow/next so the build pipeline compiles 'use workflow' / 'use step' files into durable code. (2) NEW lib/workflows/booking-reminder.ts. Per Vercel WDK conventions: workflow function (sandboxed VM, no Node.js APIs, 'use workflow' directive) does ONLY orchestration. All real work (DB queries, SMS/email sends, activity inserts) lives in step functions ('use step' directive, full Node.js access). The workflow: (a) loads booking via loadBookingForReminder step, (b) early-returns if status != 'scheduled' OR if <24h to startsAt, (c) sleeps with workflow.sleep(new Date(startsAt - 24h)) — durable, no compute consumed during the sleep, (d) re-fetches the booking after sleep (it may have been cancelled or rescheduled in the meantime; if rescheduled to >25h out we skip and let a future scheduled run pick it up), (e) calls sendReminder step which checks workspace integrations: hasTwilio → sendSmsFromApi via lib/sms/api with the contact's phone; else hasResend → sendEmailFromApi via lib/emails/api; else returns 'no_channel_configured', (f) writes an activity row via logReminderActivity step (type='reminder_sent_sms' / 'reminder_sent_email' / 'reminder_skipped' with the reason). (3) WIRED INTO submitPublicBookingAction. Right after emitSeldonEvent('booking.created', ...), the action does: const { start } = await import('workflow/api'); await start(bookingReminderWorkflow, [bookingId]). Fire-and-forget — start() returns immediately. Errors caught + logged so workflow trigger failures don't fail the booking creation itself. (4) ALONGSIDE existing workflow_runs system. workflow_runs powers /automations archetype configurator (Soul-integrated rule-based patterns); booking-reminder is a generic durable flow that doesn't need Soul. Both coexist. NO migrations. NO new env vars on the SF side, BUT requires Vercel Workflow infrastructure on the deployment side — Vercel Workflows is included in Vercel platform; no marketplace install needed. Backend redeploy required (Vercel detects the new workflow on next build via withWorkflow + lib/workflows/ scan).",
+  "description": "MCP server for SeldonFrame — AI-native Business OS platform. v1.28.3: SKILL-PACK ARCHITECTURE (CRM-side; MCP package version bumped for SemVer hygiene). Refactor with NO behavior change — same composed system prompt, different code shape. PHILOSOPHY (Karpathy / antifragile / thin-harness-fat-skill): behavioral guidance for agents (the FAT SKILL layer) belongs in markdown-style skill FILES, not inline string concatenation in prompt.ts. To improve agent intelligence, edit a skill file. No code change to the composer or runtime. As Claude (and successor models) get better at multi-turn memory, temporal reasoning, and avoiding hallucinated actions, we EDIT prose; architecture stays stable. Smaller models lean on verbose skills; larger models need terse ones. Same architecture either way. WHAT SHIPS: (1) NEW lib/agents/skills/<archetype>/<skill-id>.ts files. Each is a default-export string of markdown-shaped content. v1.28.3 ships 3 skills for the website-chatbot archetype: temporal-reasoning.ts (templated; uses {{currentDate}}, {{currentTime}}, {{timezone}} placeholders), be-smart-by-default.ts (the 7 behavioral rules introduced in v1.27.7-v1.27.8), hard-rules.ts (the 8 non-negotiable safety invariants). (2) NEW lib/agents/skills/registry.ts. Ordered REGISTRY array of {id, content, archetypes[], renderVars[]} entries. Functions: getSkillsForArchetype(archetype) returns ordered skills applicable to that archetype; renderSkill(skill, vars) substitutes {{placeholder}} tokens with values from a vars dict (unknown placeholders left intact, visible as a debug hint). (3) REFACTORED prompt.ts. composeSystemPrompt now: imports getSkillsForArchetype + renderSkill from registry; computes skillVars (currentDate, currentTime, timezone from workspace tz); emits 'up-front' skills (temporal-reasoning, be-smart-by-default) before dynamic operator content; emits 'hard-rules' AFTER dynamic content (so safety invariants are the final word the LLM reads — anchoring effect). The previous ~50 lines of inline backtick-string-concatenated prose are gone, replaced by 6 lines of registry iteration. (4) BEHAVIORAL OUTPUT IS IDENTICAL to v1.28.2. The skill files contain the EXACT same prose that was inlined; the registry orders them the same way the inline code did. Verifiable via tsc + comparing System prompts byte-for-byte. EXTENSIBILITY: adding a new behavioral skill = create skills/<archetype>/<id>.ts + add a registry entry. Zero changes to composer or runtime. Renaming an existing skill = move file + update registry + update import; prompt output identical. Voice + SMS archetypes can introduce specialized skills (skills/voice-receptionist/no-emojis.ts, skills/sms-followup-bot/under-160-chars.ts) without composer code changes. NO MCP tool changes. NO migrations. NO new env vars. Backend redeploy required.",
   "description": "MCP server for SeldonFrame — AI-native Business OS platform. v1.28.2: EMBED WIDGET POLISH (CRM-side; MCP package version bumped for SemVer hygiene). The /api/v1/public/agent/<slug>/embed.js IIFE that renders on operators' websites — first impression for end customers — was functional but plain. v1.28.2 brings it to world-class without forking a hosted-app codebase like mckaywrigley/chatbot-ui (wrong shape — that's a full Next.js app, we need a single <script> tag with zero deps on the operator's site). Polish in 5 categories: (1) MARKDOWN RENDERING. Inline parser in the IIFE handles **bold** → <strong>, *italic* → <em>, `code` → <code>, [text](https://...) → <a target=_blank rel=noopener noreferrer>, '- bullet' lines → <ul><li>, and \\n → <br>. All output goes through escapeHtml() FIRST then sanctioned tags re-injected — no raw HTML survives. SSE-friendly: re-renders innerHTML on every delta so bold/links resolve as soon as the closing token arrives. Inline CSS for strong/em/a/code/ul/li within .sf-agent-msg.assistant. (2) MOBILE-FIRST. Below 640px breakpoint the panel becomes FULL-SCREEN (Intercom pattern) — no rounded corners, fills viewport. Bubble shifts to bottom-right corner with smaller margin. Bigger touch targets throughout. (3) BRAND INHERITANCE. Header now shows the operator's logo if workspace.theme.logoUrl is set (https://... only — schema-validated server-side); falls back to a colored circle with the first letter of the org name. Logo container has bg=rgba(255,255,255,.18) so it sits cleanly on top of the operator's primaryColor header. (4) A11Y. role=dialog + aria-label on panel; role=log + aria-live=polite + aria-relevant=additions on messages region (announces new agent messages without spamming during streaming); screen-reader-only label on the textarea; aria-expanded toggles on the bubble; ARIA-labels on every control. Esc key closes panel + returns focus to bubble (focus management). focus-visible outlines on all interactive elements with proper offset. prefers-reduced-motion media query disables all transitions and animations. (5) ANIMATIONS + UX. Bubble: spring scale (cubic-bezier(.34,1.56,.64,1)) on hover (1.08x) + active (.95x). Panel: spring slide-up reveal — opacity 0 + translateY(8px) + scale(.98) → opacity 1 + 0 + 1 in 280ms. Messages: per-message fade+slide-in (220ms). Typing indicator: 3 animated bouncing dots (was 'Typing...' text). Textarea (was input) auto-grows to 120px max-height; Enter sends, Shift+Enter newline. Send button has translateY hover + scale active feedback. NO MCP tool changes. NO migrations. NO new env vars. The embed.js endpoint is automatically picked up on the next page load — operators don't need to update their script tag (the URL is stable; the response body changes server-side). Backend redeploy required.",
   "description": "MCP server for SeldonFrame — AI-native Business OS platform. v1.28.1: 90%-FEWER-TOOL-CALLS + /AGENTS DASHBOARD REDESIGN. PROBLEM: real-world dogfood showed Claude Code spending 3m 35s + ~17 actions to update an existing chatbot — only 2 of which (the actual update_agent_blueprint call + final response) were necessary work. The waste broke down as: progressive tool-schema loading (5+ separate 'Called seldonframe' round-trips as Claude Code lazy-discovered the 138-tool catalog), local-FS exploration (ls + node --version + read .env — all irrelevant on a hosted SF platform), an unnecessary question to the user about Anthropic key configuration (the workspace already had one), and stale references to the v1.27.9-removed token-budget concept leaking from MCP tool responses. Plus the /agents dashboard exposed only 2 CTAs (Open sandbox / View conversations) and zero inline health stats — operators had to drill into /agents/[id] just to see eval status / validator pass rate / 24h conversations. WHAT SHIPS: (1) NEW GET /api/v1/workspace-state ENDPOINT. Workspace-bearer auth. Returns in ONE round-trip: workspace identity (id, name, slug, industry, timezone, dashboard_url, public_site_url); integrations status (anthropic / openai / twilio / resend / kit / mailchimp configured? — booleans only, no keys leaked); agents WITH inline health stats (status, version, eval pass rate, validator pass rate 24h, conversations 24h, eval_meets_publish_gate boolean, last_eval_run_at); high-level counts (contacts, bookings, deals, agents); and a tailored next_steps[] array (e.g. 'Configure Anthropic LLM key' / 'No agents yet, call build_website_chatbot' / 'Run evals before promoting to live'). (2) NEW get_workspace_state MCP TOOL wrapping that endpoint. Description tells Claude Code to call it FIRST for any workspace task — replaces 4-6 progressive discovery calls. (3) NEW update_website_chatbot SKILL BUNDLE. Peer to v1.28.0's build_website_chatbot, for the 'agent already exists' case. Auto-resolves the workspace's website-chatbot agent_id if not passed (uses get_workspace_state internally), merges patch into blueprint via update_agent_blueprint, returns dashboard URL + version + next_steps with re-run-evals reminder. Refuses ambiguously (multi-agent workspaces) — returns the agent list instead. (4) STRIPPED DEPRECATED FIELDS. tokensUsedToday + dailyTokenBudget removed from list_agents and get_agent_metrics MCP tool response shapes (token-budget concept was retired in v1.27.9; stale data was still leaking through and causing Claude Code to suggest 'bump the token budget' as a follow-up). (5) WELCOME.JS REWRITTEN. New 'Step zero — call get_workspace_state' section + new ANTI-PATTERNS table explicitly enumerating what NOT to do (don't ls/cat/read .env; don't check node/npm versions; don't ask 'is Anthropic key configured?'; don't create duplicate agent). FIRST_CALL_BANNER similarly anchored on step-zero discovery + anti-patterns. (6) /AGENTS PAGE REDESIGN. Each agent row now shows inline: status pill + 3-stat grid (conversations 24h, validator pass rate 24h with 'X turns' hint, eval pass rate X/Y with last-run timestamp + meets-gate tone) + a contextual PromoteHint banner ('Evals haven't run' / 'Below 87.5% gate, fix in Settings' / 'Eval gate met, ready to promote') + 5 quick-link CTAs (Overview / Sandbox / Conversations / Evals / Settings — was 2). The agent list is now a roster + health dashboard, not a list-of-links. EXPECTED IMPACT: 'build a chatbot for [biz]' / 'update the chatbot's FAQ' / 'how is my chatbot doing' prompts resolve in 1-2 tool calls + ~5-10s, instead of 12+ tool calls + ~3.5min. Operators see at a glance whether each agent is healthy + what's blocking promotion, without drilling into per-agent pages. NO migrations, NO new env vars. Backend redeploy required.",
   "description": "MCP server for SeldonFrame — AI-native Business OS platform. v1.28.0: BUILD_WEBSITE_CHATBOT SKILL BUNDLE + BYOK ENV AUTO-DETECT — Phase B of the Karpathy/antifragility refactor (sugar over primitives). PROBLEM: even after v1.27.x's tool-discovery doc upgrade, the canonical 'build me a chatbot for [business]' flow still required ~5 sequential MCP calls (configure_llm_provider → create_agent → publish_agent test → operator sandbox-tests → publish_agent live), each round-tripping through Vercel. ~30s wall time. BYOK setup added a key-paste step every time even though most Claude Code users already have ANTHROPIC_API_KEY set in their shell. WHAT SHIPS: (1) BYOK ENV AUTO-DETECT. configure_llm_provider gains a 'pass nothing / pass api_key=\"env\"' mode. The MCP server reads process.env.ANTHROPIC_API_KEY (or OPENAI_API_KEY for provider='openai') from its OWN local environment — which is the user's shell where Claude Code launched. Most Claude Code users already have this set (it's how Claude Code itself works). Returns { ok: false, error: 'no_env_key' } with a clear hint if the env var isn't set. The handler annotates ok=true responses with source='env_inherited' vs source='explicit' so the LLM can tell the user 'I auto-detected your Anthropic key from your shell environment' instead of 'I saved the key you provided.' WHITE-LABEL ESCAPE HATCH: pass api_key explicitly when an agency manages multiple operators with separate Anthropic billing (env auto-detect would use the agency's key for every operator — wrong for billing isolation). (2) NEW SKILL-BUNDLE TOOL build_website_chatbot. ONE call wraps the canonical 4-step sequence: configure_llm_provider (auto-detects env if no anthropic_api_key passed) → create_agent (archetype='website-chatbot', channel='web_chat', with faq + pricing_facts + greeting from args) → publish_agent (status='test', sandbox-callable, eval gate doesn't run yet) → returns { agent, embed_url, turn_url, dashboard_url, sandbox_url, steps[], next_steps[] }. Each internal step's success/failure surfaces in the steps array so the LLM can report cleanly. Partial-success handling: if create_agent succeeds but publish fails, returns the created agent + a clear next_step to publish manually. Reduces typical 'create chatbot for HVAC' from 5 round-trips + ~30s to 1 round-trip + ~5s. (3) WELCOME.JS CAPABILITY MAP UPDATED. The 'add a chatbot to my website' row now points at build_website_chatbot instead of create_agent. Canonical short flow walkthrough rewritten to show the 1-call version. FIRST_CALL_BANNER similarly updated. Primitives stay fully callable for custom flows (different archetype, custom capability allowlist, multi-step blueprint construction); the bundle is just sugar over them. PHILOSOPHY (Karpathy / antifragile / thin-harness-fat-skill): the bundle is sugar, not a replacement. Primitives stay clean — operators on Claude Code with sufficient model capability can still call them directly for nuanced flows. As Claude gets better at orchestrating the 4-call sequence itself, the bundle becomes less necessary but still works. Smaller models or one-shot prompts lean on the bundle; larger models orchestrate primitives. Same architecture serves both. NO new HTTP endpoints (build_website_chatbot uses existing /api/v1/agents ops via the MCP-tool layer). NO migrations, NO new env vars on the server side. ENCRYPTION_KEY still required (already shipped). DEFERRED to v1.28.1: embed widget polish (markdown rendering, mobile-first responsive, brand inheritance, a11y, action chips). DEFERRED to v1.28.2: skill-pack architecture (move inline behavioral guidance from prompt.ts to agents/<archetype>/skills/*.md files the composer reads on demand). DEFERRED to v1.28.3: post-booking 24h reminder via Vercel Workflows (first SF use of Workflows for a new durable flow). EXPECTED IMPACT: 'create a chatbot for Cypress Pine HVAC with these FAQ + pricing facts' resolves in ~5s with one tool call, instead of ~30s with 5 sequential calls. Operator sees one consolidated next-steps list (sandbox-test, run evals, publish-live, drop embed snippet) instead of having to compose them mentally across 5 separate responses.",

package/src/welcome.js

@@ -8,7 +8,7 @@
 // stripped. `create_full_workspace` is the only workspace-creation
 // path mentioned anywhere in this briefing.
 
-export const VERSION = "1.28.2";
+export const VERSION = "1.28.4";
 
 export const WELCOME_MARKDOWN = `# SeldonFrame — create a real Business OS in one conversation
 
@@ -433,4 +433,4 @@ admin dashboard. Pre-fills their email automatically.
 <https://seldonframe.com> · **Discord:** <https://discord.gg/sbVUu976NW>
 `;
 
-export const FIRST_CALL_BANNER = `🚀 SeldonFrame v1.28.2 is connected. STEP ZERO: for any workspace task call get_workspace_state({workspace_id}) FIRST — returns workspace identity + integrations status + agents with inline health stats + counts + tailored next_steps in ONE round-trip. Replaces 4-6 progressive discovery calls. ANTI-PATTERNS: don't ls/cat/read .env (SF is hosted, not local files); don't check node/npm versions (irrelevant); don't ask 'is Anthropic key configured?' (state response tells you); don't create a duplicate agent (state response shows existing ones — use update_website_chatbot instead of build_website_chatbot when one exists). Token-budget concept removed in v1.27.9 — ignore stale references. CAPABILITY MAP — pick the right primitive BEFORE exploring tools: (a) "build me a website" → WORKSPACE → create_workspace_v2 + block tools. (b) "add a chatbot to my website / landing page" → AGENT (web chat) → build_website_chatbot (v1.28+ skill bundle: configure_llm + create_agent + publish-test in 1 call; auto-detects ANTHROPIC_API_KEY from env). Drop to primitives (configure_llm_provider + create_agent + publish_agent) only for custom flows. CRITICAL ANTI-PATTERN: chat widgets are NOT blocks. Don't search list_blocks for chat — chat agents are a separate primitive (v1.26+). (c) "auto-reply to inbound SMS/email" → CONVERSATION → send_conversation_turn (one-shot Soul-aware reply, not a website widget). (d) "add hero/services/faq/cta section" → BLOCK → get_block_skill + persist_block. (e) "send campaign email/sms" → MESSAGING. (f) CRM ops → list_contacts/create_deal/etc. CHATBOT CANONICAL FLOW (v1.28+ — 1 call instead of 5): build_website_chatbot({workspace_id, name, faq, pricing_facts, greeting}) → wraps configure_llm + create_agent + publish-test in one call, auto-detects ANTHROPIC_API_KEY from env. Then: operator sandbox-tests at /agents/[id]/test → publish_agent({status:"live"}) auto-runs 8-scenario eval gate (≥87.5% pass required). Observability tools after publish: list_agents, tail_agent_conversations, get_agent_conversation, get_agent_metrics, run_agent_evals, replay_conversation. Dashboard surfaces /agents, /agents/[id]/test, /agents/[id]/settings, /agents/[id]/evals, /agents/[id]/conversations let operators iterate without leaving the browser. WORKSPACE FLOW (legacy reference): create_workspace_v2 → IN PARALLEL for all 7 recommended_blocks (hero, services, about, faq, cta, booking, intake): get_block_skill + persist_block → complete_workspace_v2 → finalize_workspace({workspace_id, email}). Run blocks in PARALLEL (Promise.all) — sequential takes 60+ seconds. v1.10+ CUSTOMIZE: regenerate_block, upload_workspace_image (image_url preferred over base64). v1.11+ STRUCTURAL: get_landing_structure, move_section, delete_section. v1.12+ COMPOSITES: add_composite_section / update_composite_section — manifest ANY block from 12 low-level primitives. Skipping finalize_workspace leaves the operator with no admin login. Every URL is real. NEVER create local files.`;
+export const FIRST_CALL_BANNER = `🚀 SeldonFrame v1.28.4 is connected. STEP ZERO: for any workspace task call get_workspace_state({workspace_id}) FIRST — returns workspace identity + integrations status + agents with inline health stats + counts + tailored next_steps in ONE round-trip. Replaces 4-6 progressive discovery calls. ANTI-PATTERNS: don't ls/cat/read .env (SF is hosted, not local files); don't check node/npm versions (irrelevant); don't ask 'is Anthropic key configured?' (state response tells you); don't create a duplicate agent (state response shows existing ones — use update_website_chatbot instead of build_website_chatbot when one exists). Token-budget concept removed in v1.27.9 — ignore stale references. CAPABILITY MAP — pick the right primitive BEFORE exploring tools: (a) "build me a website" → WORKSPACE → create_workspace_v2 + block tools. (b) "add a chatbot to my website / landing page" → AGENT (web chat) → build_website_chatbot (v1.28+ skill bundle: configure_llm + create_agent + publish-test in 1 call; auto-detects ANTHROPIC_API_KEY from env). Drop to primitives (configure_llm_provider + create_agent + publish_agent) only for custom flows. CRITICAL ANTI-PATTERN: chat widgets are NOT blocks. Don't search list_blocks for chat — chat agents are a separate primitive (v1.26+). (c) "auto-reply to inbound SMS/email" → CONVERSATION → send_conversation_turn (one-shot Soul-aware reply, not a website widget). (d) "add hero/services/faq/cta section" → BLOCK → get_block_skill + persist_block. (e) "send campaign email/sms" → MESSAGING. (f) CRM ops → list_contacts/create_deal/etc. CHATBOT CANONICAL FLOW (v1.28+ — 1 call instead of 5): build_website_chatbot({workspace_id, name, faq, pricing_facts, greeting}) → wraps configure_llm + create_agent + publish-test in one call, auto-detects ANTHROPIC_API_KEY from env. Then: operator sandbox-tests at /agents/[id]/test → publish_agent({status:"live"}) auto-runs 8-scenario eval gate (≥87.5% pass required). Observability tools after publish: list_agents, tail_agent_conversations, get_agent_conversation, get_agent_metrics, run_agent_evals, replay_conversation. Dashboard surfaces /agents, /agents/[id]/test, /agents/[id]/settings, /agents/[id]/evals, /agents/[id]/conversations let operators iterate without leaving the browser. WORKSPACE FLOW (legacy reference): create_workspace_v2 → IN PARALLEL for all 7 recommended_blocks (hero, services, about, faq, cta, booking, intake): get_block_skill + persist_block → complete_workspace_v2 → finalize_workspace({workspace_id, email}). Run blocks in PARALLEL (Promise.all) — sequential takes 60+ seconds. v1.10+ CUSTOMIZE: regenerate_block, upload_workspace_image (image_url preferred over base64). v1.11+ STRUCTURAL: get_landing_structure, move_section, delete_section. v1.12+ COMPOSITES: add_composite_section / update_composite_section — manifest ANY block from 12 low-level primitives. Skipping finalize_workspace leaves the operator with no admin login. Every URL is real. NEVER create local files.`;