leedab 0.5.3 → 0.6.1

package/README.md

@@ -1,6 +1,6 @@
 # LeedAB
 
-**The AI operating system for companies. A brain that learns your business, Artificial Beings that run it.**
+**The AI operating system for logistics. A brain that learns how your business operates, Artificial Beings that execute on the apps your team already use.**
 
 LeedAB installs on any Mac, Windows, or Linux computer and becomes the operating system for your back-office. A shared **brain** ingests Slack, email, sheets, tickets, app screens, and docs, then learns how your business actually runs: how a missed delivery gets recovered, how a credit note gets drafted, who owns the priority shipments, the edge cases that used to live in operators' heads. Specialist **Artificial Beings (ABs)** drive the apps your team already uses, carrier portals, TMS, ERPs, inboxes, spreadsheets, the same way a human operator does. No API connectors. No IT project. Local-first. Every byte of data stays on your hardware.
 
@@ -67,4 +67,9 @@ The one-liner installer handles Node.js and all dependencies for you.
 
 ## License
 
-Proprietary. See [LICENSE](./LICENSE) for details.
+LeedAB is commercial software, all rights reserved. Installing this
+package from npm does not grant a right to use it — a paid license key
+is required for the runtime to function. Visit
+[leedab.com](https://www.leedab.com) to obtain one.
+
+Full terms: [leedab.com/terms](https://www.leedab.com/terms).

package/_schema/customer-brain-template/01_Customers/engagement.md

@@ -0,0 +1,42 @@
+---
+type: engagement-plan
+client_id: {{CLIENT_ID}}
+display_name: {{DISPLAY_NAME}}
+industry: {{INDUSTRY}}
+status: active
+phase: kickoff
+day_of_program: 0
+current_focus: null
+next_milestone_iso: null
+champion: null
+provisioned_at: {{PROVISIONED_DATE}}
+last_updated: {{PROVISIONED_DATE}}
+---
+
+# {{DISPLAY_NAME}} — engagement plan
+
+> **Owned by Onboarding.** Champion reads this via `read_engagement` for grounding on every interaction.
+> **Updates:** POST `/api/onboarding/engagement-plan` with `{ clientId, patch?, bodyAppend? }`.
+> **Initialised:** {{PROVISIONED_DATE}}.
+
+## Current state
+
+(Onboarding fills this in as the customer moves through Day-0 → Day-7 → Day-30 → Day-90.)
+
+## Phase milestones
+
+| Day | Milestone | Owner | Status |
+|---|---|---|---|
+| 0 | Kickoff brief reviewed + Champion designated | Onboarding | pending |
+| 7 | First win replaced — Champion confirms | Onboarding | pending |
+| 30 | Health check — sticking? | Onboarding + CX | pending |
+| 60 | Expansion conversation — what else can we automate? | Delivery | pending |
+| 90 | Renewal-readiness QBR | Delivery | pending |
+
+## Recent decisions
+
+(Append milestone notes here via the engagement-plan endpoint's `bodyAppend`.)
+
+## Open blockers
+
+(Things stalling the engagement — flagged here so Champion can route them up.)

package/_schema/customer-brain-template/01_Customers/qbr-template.md

@@ -0,0 +1,49 @@
+---
+type: qbr-template
+client_id: {{CLIENT_ID}}
+display_name: {{DISPLAY_NAME}}
+quarter: ""
+prepared_by: Delivery agent
+delivered_iso: null
+status: draft
+---
+
+# {{DISPLAY_NAME}} — Q? QBR
+
+> **Owned by Delivery agent.** Generated by the `qbr_brief` tool. Reviewed by Champion before delivery. Goes to {{DISPLAY_NAME}}'s sponsor (usually the COO / GM).
+
+## Health signal
+
+**(GREEN / AMBER / RED)** — *(Delivery sets this. Tied to engagement-plan health checks + CX scores.)*
+
+## By the numbers
+
+| Metric | This quarter | Last quarter | Direction |
+|---|---|---|---|
+| Hours saved (estimated) | *(0)* | *(0)* | *(↑/↓/→)* |
+| Successful agent runs | *(0)* | *(0)* | *(↑/↓/→)* |
+| Triaged inbound (CX) | *(0)* | *(0)* | *(↑/↓/→)* |
+| Health score (avg) | *(0)* | *(0)* | *(↑/↓/→)* |
+| MRR (if tier-priced) | *(0)* | *(0)* | *(↑/↓/→)* |
+
+## Highlights to lead with
+
+(Two or three concrete wins. Power-user adoption stories. Saved-revenue events. Things the customer's sponsor will want to repeat at their own internal review.)
+
+## Risks to surface
+
+(Two or three concrete things that could degrade the engagement if unaddressed. Be honest. Champion is in the room — surfaces are easier to fix when named.)
+
+## Asks for them
+
+(What we want from {{DISPLAY_NAME}}. e.g. reference call, case-study consent, intro to a similar prospect, expanded license.)
+
+## Next quarter focus
+
+(One sentence. What's the headline outcome we're chasing together?)
+
+## Champion brief (pre-meeting)
+
+> Send to Champion ahead of the QBR call. They'll walk the sponsor through with this as their reference.
+
+(The Champion-side talking points. Lighter than the QBR itself — bullet-form is fine.)

package/_schema/customer-brain-template/02_Sales/pipeline.md

@@ -0,0 +1,43 @@
+---
+type: sales-pipeline
+client_id: {{CLIENT_ID}}
+display_name: {{DISPLAY_NAME}}
+last_updated_iso: {{PROVISIONED_DATE}}
+total_face_value: 0
+total_weighted_value: 0
+deal_count: 0
+---
+
+# {{DISPLAY_NAME}} — sales pipeline
+
+> **Owned by Sales agent.** Read via `pipeline_summary` tool; updated by the Sales agent or operator (manual edit + agent-driven append).
+>
+> **Stages are customer-defined.** Edit the columns below to match {{DISPLAY_NAME}}'s actual sales motion. Don't force a SaaS template onto a logistics workflow or vice versa.
+
+## Active deals
+
+| Deal | Stage | Face value | Weighted (% × face) | Owner | Last touched | Next action |
+|---|---|---|---|---|---|---|
+| (e.g. Acme Logistics — SYD-MEL contract) | Discovery | $18,000 | $5,400 (30%) | Sales | 2026-MM-DD | Demo Tue 14:00 |
+| ... | ... | ... | ... | ... | ... | ... |
+
+## Stages (configure for {{DISPLAY_NAME}})
+
+| Stage | Default close-prob | Definition |
+|---|---|---|
+| Discovery | 30% | Customer signals intent. Need + budget unconfirmed. |
+| Demo'd | 50% | They've seen the workflow live. |
+| Proposal | 70% | They have a quote. Reading it. |
+| Verbal | 85% | They've said yes verbally. Contract pending. |
+| Won | 100% | Signed. |
+| Lost | 0% | Stalled or chose someone else. Capture the why in `won-lost-postmortems/`. |
+
+## This week's focus
+
+(One or two named deals. Updated weekly by Sales. Keep this short — the pipeline is the table above; this is a focus pointer.)
+
+## Recent moves
+
+(Append-only log of stage transitions. Sales agent appends via the `pipeline_summary` tool; operators can edit manually.)
+
+- 2026-MM-DD — *(Deal X moved Discovery → Demo'd. Reason: …)*

package/_schema/customer-brain-template/04_Operations/runbook.md

@@ -0,0 +1,54 @@
+---
+type: runbook
+client_id: {{CLIENT_ID}}
+title: ""
+owner: ""
+status: draft
+last_verified_iso: null
+last_run_iso: null
+---
+
+# {{DISPLAY_NAME}} runbook — *(rename this file to the runbook's slug)*
+
+> **What this is.** A reusable playbook for one specific repeatable task. The canonical place an agent looks when they need to actually DO this thing rather than reason about it from first principles every time.
+>
+> **Who owns it.** One person at {{DISPLAY_NAME}} is the owner — they're accountable for keeping it accurate. Agents propose updates; the owner reviews + accepts.
+
+## When this fires
+
+(Trigger condition. e.g. "First Tuesday of every month" / "Whenever a Coles deduction lands in the inbox" / "When Champion flags a Day 30 health drop.")
+
+## Inputs required
+
+(Concrete data the agent or operator needs in hand before starting. e.g. "Last 30 days of dispatch records" / "Latest Stripe export" / "Champion's most recent confidence score.")
+
+## Steps
+
+1. **Step 1.** What happens. Who/what does it. Expected duration.
+2. **Step 2.** ...
+3. **Step 3.** ...
+
+(Keep steps imperative + atomic. If a step has sub-decisions, break it out.)
+
+## Outputs / artefacts
+
+- (e.g. "Variance report posted to Champion's Telegram thread")
+- (e.g. "Updated `02_Sales/pipeline.md` with new weighted value")
+
+## Failure modes + recovery
+
+| What can go wrong | How you'd notice | Recovery |
+|---|---|---|
+| (e.g. SAP order ID mismatch) | (e.g. validation step fails) | (e.g. flag to Ops manager via Telegram, hold dispatch) |
+
+## Last verified
+
+- **Date:** *(the runbook owner sets this when they walk through the steps end-to-end and confirm they still match reality)*
+- **Verified by:** *(name)*
+- **Notes:** *(what changed since the last verify, if anything)*
+
+## Cross-references
+
+- (Link to related runbooks)
+- (Link to ADRs that constrain this runbook's scope)
+- (Link to customer-side wiki pages for context)

package/_schema/customer-brain-template/05_Finance/runway-snapshot.md

@@ -0,0 +1,40 @@
+---
+type: runway-snapshot
+client_id: {{CLIENT_ID}}
+display_name: {{DISPLAY_NAME}}
+last_snapshot_iso: {{PROVISIONED_DATE}}
+cash_on_hand: null
+monthly_burn: null
+mrr: null
+runway_months: null
+---
+
+# {{DISPLAY_NAME}} — runway snapshot
+
+> **Owned by Finance agent.** Generated by the `runway_snapshot` tool. Updated weekly (or on demand). The numbers below are placeholders until Finance runs its first read.
+
+## Current state
+
+| Metric | Value |
+|---|---|
+| Cash on hand | *(unset)* |
+| Monthly burn (3mo avg) | *(unset)* |
+| MRR | *(unset)* |
+| Net burn (burn − MRR) | *(unset)* |
+| **Runway** | *(unset)* |
+
+## Notes
+
+(Plain-English context for the numbers. Why burn moved, what's coming up that'll change it. Append-only.)
+
+## Levers if needed
+
+(Specific moves Finance + the customer team have pre-agreed for cash crunches. Editable; review quarterly.)
+
+1. *(e.g. Pause Q4 hiring — saves $X/mo)*
+2. *(e.g. Renegotiate vendor Y — saves $Y/mo)*
+3. *(e.g. Outbound focus: 2 more $5k MRR deals push net-burn negative)*
+
+## Watch
+
+(What Finance is keeping an eye on for next snapshot — e.g. invoices due, large vendor renewals, anomalous spend.)

package/_schema/customer-brain-template/CLAUDE.md

@@ -0,0 +1,66 @@
+# {{DISPLAY_NAME}} Brain — vault constitution
+
+This vault is **{{DISPLAY_NAME}}'s nervous system** running on LeedAB. LLM-maintained operational memory shared between {{DISPLAY_NAME}} and their LeedAB agent layer.
+
+## What this vault is for
+
+LeedAB is an AI co-founder system for SMBs. This vault holds {{DISPLAY_NAME}}'s organisational memory: customer records, contracts, team, decisions, dispatches, sales pipeline, financials, ops runbooks, product info, legal templates.
+
+**Audiences:**
+- {{DISPLAY_NAME}} team (full read/write through the dashboard + their Champion)
+- LeedAB agents (read + write via authenticated processes per the manifest)
+- LeedAB founders (Mina, Muiez, Ziad) for support / onboarding only
+
+## Security-first — load-bearing
+
+**Default to private.** Customer data, financials, decisions, and agent context all live here.
+
+- `_secrets/` NEVER syncs and NEVER holds credentials. Read-only references to where credentials live (1Password / OS Keychain).
+- Per-customer isolation: this brain only contains {{DISPLAY_NAME}}'s data. Cross-customer references forbidden.
+- Audit content before any external action.
+- Threat-model before publishing anything.
+
+## Architecture — adapted PARA + business categories
+
+- `00_Org/` — team, advisors, partners, vendors
+- `01_Customers/` — {{DISPLAY_NAME}}'s own customers
+- `02_Sales/` — pipeline, prospects, demos
+- `03_Product/` — roadmap, decisions, ADRs
+- `04_Operations/` — SOPs, runbooks, postmortems
+- `05_Finance/` — revenue, costs, runway
+- `06_Legal/` — contracts, IP, compliance, templates
+- `Permanent/` — atomic claims (Zettelkasten-style)
+- `Wiki/` — topic wikis (industry knowledge, customer use cases, agent design patterns)
+- `_inbox/` — daily notes + agent captures + auto-captured sessions
+- `_templates/` — note templates
+- `_attachments/` — PDFs, images, design files
+- `_secrets/` — local-only credential references (NEVER syncs)
+
+## Operational rules
+
+- `_secrets/` never syncs (verify on every device)
+- `.obsidian/workspace.json` excluded per device
+- Templates enforced — every customer note from `_templates/customer.md`, decisions use `_templates/decision.md`
+- Idempotent agent writes (re-running same input → same output, no duplicate appends)
+- One canonical writer per file (agents serialise writes)
+- Cross-customer isolation strict
+
+## Manifest
+
+This vault is bound to:
+- **Client ID:** {{CLIENT_ID}}
+- **Industry:** {{INDUSTRY}}
+- **Tier:** {{TIER}} (or null for prospect / non-tiered installs)
+- **Manifest:** `LeedAB-OS/clients/{{CLIENT_ID}}/manifest.json`
+
+## What NOT to do
+
+- Don't put credentials in this vault
+- Don't reference another customer's data
+- Don't echo credentials in chat or logs
+- Don't sync `_secrets/`
+- Don't fabricate claims — empty pages > invented ones
+
+## Build state
+
+This brain was provisioned by `bin/onboard-customer.sh` on {{PROVISIONED_DATE}}. Champion onboarding flow: ask {{DISPLAY_NAME}}'s designated Champion to walk through the Day-0 brief in `_inbox/champion-day-0.md`.

package/_schema/customer-brain-template/README.md

@@ -0,0 +1,43 @@
+# {{DISPLAY_NAME}} Brain
+
+> Private brain for {{DISPLAY_NAME}} ({{INDUSTRY}}). Provisioned {{PROVISIONED_DATE}}.
+
+Don't share this URL externally. Don't push to a public repo.
+
+For Champion onboarding, see `_inbox/champion-day-0.md`.
+
+## Bound configuration
+
+- **Client ID:** {{CLIENT_ID}}
+- **Industry:** {{INDUSTRY}}
+- **Tier:** {{TIER}}
+- **Manifest:** `LeedAB-OS/clients/{{CLIENT_ID}}/manifest.json`
+- **Playbook:** `LeedAB-OS/context/customers/{{CLIENT_ID}}/playbook.md`
+
+## Folders
+
+| Folder | Purpose |
+|---|---|
+| `00_Org/` | Team, advisors, vendors |
+| `01_Customers/` | {{DISPLAY_NAME}}'s own customers |
+| `02_Sales/` | Pipeline + prospects |
+| `03_Product/` | Roadmap, ADRs |
+| `04_Operations/` | SOPs, runbooks |
+| `05_Finance/` | Revenue, runway |
+| `06_Legal/` | Contracts, compliance |
+| `Permanent/` | Atomic claims |
+| `Wiki/` | Topic wikis |
+| `_inbox/` | Daily notes + agent captures |
+| `_templates/` | Note templates |
+| `_attachments/` | PDFs, images |
+| `_secrets/` | LOCAL-ONLY credential refs |
+
+## Sync
+
+This brain is meant to live on the customer's substrate (Mac Mini today, UGREEN NAS appliance later). Backed by Obsidian Sync subscription. `_secrets/` excluded by default.
+
+## Cross-references
+
+- LeedAB OS: https://github.com/muiez/LeedAB-OS
+- Champion contact: TBD (set during onboarding call)
+- Support: contact your LeedAB Champion via Telegram

package/_schema/customer-brain-template/_inbox/champion-day-0.md

@@ -0,0 +1,69 @@
+---
+type: champion-onboarding
+client_id: {{CLIENT_ID}}
+provisioned_at: {{PROVISIONED_DATE}}
+status: pending
+---
+
+# Champion onboarding — {{DISPLAY_NAME}} (Day 0)
+
+> Hi! You're the designated **Champion** for LeedAB at {{DISPLAY_NAME}}. This page walks you through Day 0 — the first hour of getting LeedAB usefully running on your team's behalf.
+
+## What you're being asked to do
+
+1. **Walk through the agent rail** — see who's on your roster.
+2. **Test a simple workflow** — message one agent, get a useful reply.
+3. **Identify your day-1 use case** — pick one task you'd love to never do again.
+4. **Set comms preferences** — when should agents reach you, and how.
+
+## Day 0 checklist (~30 minutes)
+
+### 1. Open the dashboard
+
+URL: `http://127.0.0.1:3100` (on the LeedAB-bound Mac Mini at your office) or your assigned hosted URL.
+
+You should see:
+- A rail on the left showing your licensed agents (per the manifest)
+- The orb view in the centre — your "workforce" visualised
+- A pack badge top-right showing your tier ({{TIER}})
+
+If anything's missing → ping your LeedAB Champion agent on Telegram immediately.
+
+### 2. Try a workflow
+
+Message any agent with a small ask. Examples:
+- COO → "What does my morning look like?"
+- CX → "Draft a reply to a customer asking about delivery delays."
+- Finance → "Show me last month's runway summary."
+
+You should get a reply within ~10-30 seconds.
+
+### 3. Identify your day-1 task
+
+Tell your Champion agent: "The thing I'd love to stop doing is {X}." That gets logged here for me to escalate to Onboarding + Implementation.
+
+### 4. Set comms preferences
+
+Update `00_Org/champion-profile.md` with:
+- Preferred Telegram window (e.g. 08:00-18:00 Sydney)
+- Voice OK or only async?
+- Tech-fluency self-rating (novice / familiar / fluent / expert)
+
+## Day 7 check-in
+
+I'll DM you on Day 7 with a structured check-in. Goal by then: at least one task replaced.
+
+## Day 30 check-in
+
+By Day 30 we want you saying "I'd be sad to lose this." If we're not there, we adjust scope or tone.
+
+## Escalation
+
+For things I can't handle:
+- Custom integrations / data migrations → I escalate to **Implementation**
+- Team-wide milestones / scope expansion → I escalate to **Onboarding**
+- Anything else (network / hardware / billing) → I escalate to **Delivery**
+
+You never need to bother Mina or Muiez directly — that's what I'm for.
+
+— Champion (your LeedAB enablement coach)

package/_schema/customer-brain-template/_inbox/weekly-digest-template.md

@@ -0,0 +1,53 @@
+---
+type: weekly-digest-template
+client_id: {{CLIENT_ID}}
+display_name: {{DISPLAY_NAME}}
+delivered_via: Telegram
+cadence: Monday 09:00 local
+recipient: Champion + sponsor
+---
+
+# {{DISPLAY_NAME}} — weekly digest
+
+> **Owned by COO orchestrator (or Delivery for non-COO installs).** Auto-generated every Monday morning. Goes to Champion + sponsor via Telegram. Edit this template to change what shows up; the digest tool reads structure from here.
+
+## Format
+
+The digest follows this exact shape so it's parseable + scannable:
+
+---
+
+### {{display_name}} — week of *(Monday date)*
+
+**On track**
+- *(2-3 wins from last week. Concrete: numbers, dollar amounts, named decisions.)*
+
+**Watch**
+- *(2-3 things drifting. Not panic-level. Naming them keeps them named.)*
+
+**Off track / blocked**
+- *(0-2 things actually broken. Each one names an owner + a next move.)*
+
+**Coming up this week**
+- *(2-3 known scheduled items. Meetings, deadlines, milestones.)*
+
+**Asks**
+- *(0-1 thing we need from Champion or sponsor this week. If nothing, omit the section.)*
+
+---
+
+## Configuration
+
+| Setting | Default | Override notes |
+|---|---|---|
+| Send time | Monday 09:00 *(Australia/Sydney)* | Set per Champion's stated preference |
+| Channel | Telegram (Champion's primary) | Add WhatsApp / email mirror if Champion requests |
+| Recipients | Champion + sponsor | Add others via Routing rules in Admin |
+| Length cap | ~250 words | Hard cap so it stays scannable on phone |
+
+## Tone
+
+- **Professional, terse.** Operator-to-operator voice, not a marketing tone.
+- **Concrete > vague.** "$42k weighted pipeline (+18%)" not "pipeline is healthy."
+- **Honest about misses.** "Off track" sections build trust; sanitised digests erode it.
+- **No internal LeedAB jargon.** Don't reference agent IDs unless it materially helps the reader.

package/_schema/customer-brain-template/_templates/customer.md

@@ -0,0 +1,33 @@
+---
+type: customer
+status: prospect | trial | active | churned
+created: {{date}}
+industry:
+size:
+primary_contact:
+billing_terms:
+---
+
+# {{customer_name}}
+
+## Snapshot
+
+- **Status:**
+- **Stage:**
+- **Onboarded:**
+
+## Contacts
+
+| Name | Role | Email | Phone | Comms preference |
+|------|------|-------|-------|------------------|
+|      |      |       |       |                  |
+
+## Engagement
+
+- 30/60/90 plan: link to `01_Customers/<name>/30-60-90.md`
+- Active workflows:
+- Open issues:
+
+## History
+
+(append-only log)

package/_schema/customer-brain-template/_templates/decision.md

@@ -0,0 +1,30 @@
+---
+type: decision
+adr_number:
+status: proposed | accepted | superseded | deprecated
+created: {{date}}
+deciders:
+supersedes:
+superseded_by:
+---
+
+# {{decision_title}}
+
+## Context
+
+What's the situation? What forces are at play?
+
+## Decision
+
+What did we decide?
+
+## Consequences
+
+- Positive:
+- Negative:
+- Neutral:
+
+## Alternatives considered
+
+- Option A — pros, cons, why not
+- Option B — pros, cons, why not