tools-only Nucleus Launch Package v1
Status: STRATEGIC DECISION
Date: January 30, 2026
Decision Authority: Founder
The Core Question
"What is the correct toolset to package for our first launch?"
The Answer
BOTH verifiability AND product-market fit, combined:
- PMF is the SELECTOR — What solves real problems?
- Verifiability is the FILTER — What works predictably?
- "One Story" is the PACKAGE — What's memorable?
The Decision Framework: "Verified Value"
┌─────────────────────────────────────────────────────────────────┐
│ LAUNCH PACKAGE DECISION │
├─────────────────────────────────────────────────────────────────┤
│ │
│ 135 Tools │
│ │ │
│ ▼ │
│ ┌─────────────────┐ │
│ │ PMF FILTER │ "Does it solve a real problem?" │
│ │ (Value-First) │ │
│ └────────┬────────┘ │
│ │ │
│ ▼ (~40 tools) │
│ ┌─────────────────┐ │
│ │ VERIFY FILTER │ "Does it work in all IDEs?" │
│ │ (Reliability) │ │
│ └────────┬────────┘ │
│ │ │
│ ▼ (~20 tools) │
│ ┌─────────────────┐ │
│ │ STORY FILTER │ "Can I demo in 60 seconds?" │
│ │ (Memorable) │ │
│ └────────┬────────┘ │
│ │ │
│ ▼ (5-7 tools) │
│ ┌─────────────────┐ │
│ │ LAUNCH PACKAGE │ "Govern Your Agents in 60 Seconds" │
│ └─────────────────┘ │
│ │
└─────────────────────────────────────────────────────────────────┘
The Launch Story: "Govern Your Agents in 60 Seconds"
Primary Narrative (5 tools)
| Step | Tool | What It Does | Demo Time |
|---|---|---|---|
| 1 | brain_mount_server |
Mount an external MCP server | 10s |
| 2 | brain_governance_status |
Show it's sandboxed (Default-Deny) | 10s |
| 3 | brain_write_engram |
Remember a decision persistently | 15s |
| 4 | brain_query_engrams |
Recall it in a new session | 15s |
| 5 | brain_audit_log |
Show the cryptographic audit trail | 10s |
Total: 60 seconds, one story, five tools.
Supporting Cast (15 tools for docs)
| Category | Tools | Purpose |
|---|---|---|
| Mounting | brain_unmount_server, brain_list_mounted, brain_discover_mounted_tools |
Recursive Aggregator |
| Memory | brain_search_memory, brain_read_memory |
Persistent context |
| Tasks | brain_add_task, brain_list_tasks, brain_get_next_task, brain_update_task |
Workflow |
| Observability | brain_health, brain_version, brain_dashboard |
Status |
| DSoR (v0.6.0) | brain_dsor_status, brain_list_decisions |
Decision provenance |
Tool Tiers (Post-Launch)
Tier 1: Free (Launch Package)
- 20 tools as defined above
- Unlimited mounts
- Local .brain/ storage
- Community support
Tier 2: Pro ($29/month)
- All Free features +
- Token metering (
brain_metering_summary) - IPC security (
brain_ipc_tokens) - Decision provenance (
brain_list_decisions) - Email support
Tier 3: Enterprise (Custom)
- All Pro features +
- Federation (
brain_federation_*) - SSO/RBAC (roadmap)
- SLA + priority support
Verification Matrix (To Complete)
| Tool | Claude Desktop | Windsurf | Antigravity | Unit Tests | Launch? |
|---|---|---|---|---|---|
brain_mount_server |
⬜ | ⬜ | ⬜ | ✅ | ⬜ |
brain_governance_status |
⬜ | ⬜ | ⬜ | ✅ | ⬜ |
brain_write_engram |
⬜ | ⬜ | ⬜ | ✅ | ⬜ |
brain_query_engrams |
⬜ | ⬜ | ⬜ | ✅ | ⬜ |
brain_audit_log |
⬜ | ⬜ | ⬜ | ✅ | ⬜ |
Action: Manually verify each tool in each IDE. Only tools with 4/4 ✅ launch.
Other Strategic Questions (Answered)
Q1: Pricing model?
A: Freemium with metered Pro tier. Free for adoption, Pro for revenue.
Q2: Which IDE first?
A: Claude Desktop (Anthropic owns MCP), then Windsurf.
Q3: How to handle 135 tools?
A: Tool Router Pattern. Group into ~10 capabilities, not 135 tools.
Q4: GTM motion?
A: Developer-first PLG. Reddit/HN → PyPI → Influencers → Discord.
Q5: vs CLAUDE.md differentiation?
A: "Context vs Control" — CLAUDE.md is static context, Nucleus is dynamic control.
Q6: Key metric?
A: Active Mounts (weekly). Not downloads, not stars.
Q7: Open source strategy?
A: Open Core. Core 20 tools open, metering/federation proprietary.
Q8: vs LangGraph/CrewAI?
A: "Protocol layer, not framework." Complementary, not competitive.
Q9: Security disclosure?
A: CVE + transparent disclosure. Already doing this (CVE-2026-001).
Q10: Minimum team?
A: Solo founder + AI for v1. First hire: community manager at 1000 users.
Q11: Wait for MCP 2.0?
A: Ship now. First mover > perfect timing.
Q12: Biggest risk?
A: Complexity paralysis. Solution: One story, 5 tools, 60 seconds.
Immediate Actions
- [ ] Complete verification matrix — Test 5 core tools in all IDEs
- [ ] Record 60-second demo video — The launch story
- [ ] Write landing page copy — "Govern Your Agents in 60 Seconds"
- [ ] Prepare Reddit/HN post — Authority framing, not hype
- [ ] Set up metering for Pro tier — v0.6.0 DSoR enables this
The Final Word
"135 tools is a moat. 5 tools is a launch."
Ship the story, not the platform. The other 130 tools are there for power users to discover. But the launch is ONE STORY, FIVE TOOLS, SIXTY SECONDS.
Strategic decision recorded January 30, 2026.