create-dss-project 0.1.0

package/template/docs/executive-summary.md

@@ -0,0 +1,104 @@
+# Executive Summary — {{PROJECT_NAME}}
+
+**Last updated**: {{DATE}}
+**Confidence**: LOW (pre-discovery)
+
+---
+
+## 1. Executive Summary
+
+**Thesis**: {{STRATEGIC_HYPOTHESIS}}
+
+**Why now**: <!-- What market shift, technology change, or competitive gap creates the opportunity now? -->
+
+**Market size**: <!-- TAM/SAM/SOM headline. Reference: `research/market/` -->
+
+**Ask**: <!-- What resources/investment/time is needed? -->
+
+**Leading hypothesis**: <!-- The current best bet for how to win -->
+
+**Verdict**: <!-- PROCEED / PIVOT / KILL — with one-sentence rationale -->
+
+---
+
+## 2. Market Opportunity
+
+### Market Sizing
+
+| Metric | Value | Methodology | Evidence Grade |
+|--------|-------|-------------|---------------|
+| TAM | | | |
+| SAM | | | |
+| SOM (Y1) | | | |
+
+### Market Dynamics
+
+**Tailwinds** (forces making this easier):
+1.
+
+**Headwinds** (forces making this harder):
+1.
+
+---
+
+## 3. Competitive Landscape
+
+<!-- Summary of competitive positioning. Full teardowns in research/competitors/ -->
+
+| Competitor | Category | Overlap | Key Differentiator | Threat Level |
+|-----------|----------|---------|-------------------|-------------|
+
+**White space**: <!-- What gap exists that no competitor fills? -->
+
+---
+
+## 4. Product Strategy
+
+**Entry wedge**: <!-- The first thing you build/sell -->
+
+**Expansion path**: <!-- What comes after the wedge succeeds -->
+
+**Key architectural decisions**: <!-- Reference memory/decisions.md -->
+
+---
+
+## 5. Go-to-Market
+
+**ICP**: {{ICP_DESCRIPTION}}
+
+**Channels**: <!-- How you reach them -->
+
+**Pricing**: <!-- Model and target price point -->
+
+**Sales motion**: <!-- PLG, sales-assisted, enterprise, founder-led -->
+
+---
+
+## 6. Kill Conditions & Discovery Progress
+
+| KC | Condition | Status | Key Evidence |
+|----|-----------|--------|-------------|
+| 1 | | UNTESTED | |
+| 2 | | UNTESTED | |
+| 3 | | UNTESTED | |
+| 4 | | UNTESTED | |
+
+**Discovery progress**: 0 calls completed. See `memory/discovery.md` for detailed tracking.
+
+---
+
+## 7. Risk Register
+
+| # | Risk | Probability | Impact | Mitigation |
+|---|------|------------|--------|-----------|
+
+---
+
+## 8. Team & Timeline
+
+**Team**: {{TEAM}}
+
+**Key milestones**:
+
+| Milestone | Target Date | Status |
+|-----------|------------|--------|

package/template/docs/getting-started.md

@@ -0,0 +1,274 @@
+# Getting Started — Your First Project
+
+This guide walks you through everything from scratch. No coding experience needed.
+
+---
+
+## What Is This?
+
+The **DS Strategy Stack** is a pre-built workspace that helps you run corporate projects — market research, competitor analysis, internal implementations, vendor evaluations, due diligence, or any project that needs structured thinking and decision-making.
+
+It works with **Claude Code**, an AI assistant that lives in your terminal (the text-based interface on your computer). You talk to Claude, and it helps you:
+- Organise your research into structured files
+- Track prospects and competitors
+- Make evidence-based decisions
+- Generate a live dashboard showing your project's progress
+
+Think of it like a strategy consulting toolkit that an AI helps you run.
+
+---
+
+## What You'll Need
+
+Before starting, you'll need three things installed on your computer:
+
+1. **Git** — a tool for downloading and tracking changes to projects
+2. **Node.js** — a tool that powers the dashboard (you won't need to write code)
+3. **Claude Code** — the AI assistant that runs the project
+
+Don't worry if you don't have these yet — we'll install them step by step.
+
+---
+
+## Step 1: Install Git
+
+### On Mac
+1. Open **Terminal** (press Cmd + Space, type "Terminal", press Enter)
+2. Type `git --version` and press Enter
+3. If Git isn't installed, your Mac will prompt you to install it — follow the prompts
+4. If it shows a version number (like `git version 2.39.0`), you're good
+
+### On Windows
+1. Go to https://git-scm.com/download/win
+2. Download and run the installer
+3. Accept all the default settings
+4. Open **Command Prompt** (press Win + R, type "cmd", press Enter)
+5. Type `git --version` to confirm it's installed
+
+---
+
+## Step 2: Install Node.js
+
+1. Go to https://nodejs.org
+2. Download the **LTS** version (the one that says "Recommended for most users")
+3. Run the installer and accept all defaults
+4. Open Terminal (Mac) or Command Prompt (Windows)
+5. Type `node --version` — you should see something like `v18.17.0`
+
+---
+
+## Step 3: Install Claude Code
+
+Claude Code is Anthropic's AI assistant for your terminal.
+
+1. Go to https://claude.ai/code and follow the installation instructions for your platform
+2. Once installed, type `claude` in your terminal to verify it works
+3. You'll need an Anthropic account — create one if you don't have one
+
+---
+
+## Step 4: Create a GitHub Account (if you don't have one)
+
+GitHub is where your project lives online. You'll need it for:
+- Storing your project
+- Deploying your dashboard (optional)
+- Collaborating with others (optional)
+
+1. Go to https://github.com and click "Sign up"
+2. Follow the prompts to create your account
+3. Choose the free plan — it has everything you need
+
+---
+
+## Step 5: Get the Template
+
+1. Open Terminal
+2. Navigate to where you want your project (e.g., `cd ~/Documents`)
+3. Run this command:
+   ```
+   git clone https://github.com/DiffTheEnder/DSS-Claude-Stack.git my-strategy-project
+   ```
+   (Replace `my-strategy-project` with whatever you want to call your folder)
+4. Move into the folder:
+   ```
+   cd my-strategy-project
+   ```
+
+---
+
+## Step 6: Set Up Your Project
+
+1. In your terminal (inside the project folder), start Claude:
+   ```
+   claude
+   ```
+2. Tell Claude to run the setup:
+   ```
+   /onboard
+   ```
+3. Claude will ask you questions about your project:
+   - **Quick Start** (5 minutes): Just the basics — project name, what it's about, your hypothesis
+   - **Full Setup** (15 minutes): Everything including scoring dimensions, kill conditions, and feature selection
+
+4. Claude will ask what type of project you're running — market entry, growth strategy, competitor research, product launch, internal implementation, vendor evaluation, due diligence, business case, transformation, or custom. Each type comes with sensible defaults for which features to enable.
+
+5. Claude will configure all the files based on your answers
+
+**Tip**: If you're new to this, choose **Quick Start** and select **"Learning"** as your goal.
+
+---
+
+## Step 7: Deploy Your Dashboard (Optional)
+
+Your project includes a live dashboard that shows your research, pipeline, and strategic progress as a website.
+
+**Important:** The dashboard is a simple starting point — think of it as a 101-level view. Every project is different, so you should customise it as you go. Add new pages, create new data flows from your research to the dashboard, or build visualisations that make sense for your specific project. You can ask Claude to help — for example, "add a new dashboard page that shows our market sizing data" or "create a chart comparing vendor scores." The dashboard is yours to extend.
+
+Here's how to put it online:
+
+### Create a Vercel Account
+1. Go to https://vercel.com
+2. Click "Sign Up" and choose "Continue with GitHub"
+3. Authorise Vercel to access your GitHub account
+
+### Push Your Project to GitHub
+1. Go to https://github.com/new to create a new repository
+2. Name it (e.g., `my-strategy-project`)
+3. Choose **Private** (your strategy data should stay confidential)
+4. Don't add a README (you already have one)
+5. Follow the "push an existing repository" instructions GitHub shows you:
+   ```
+   git remote set-url origin https://github.com/YOUR-USERNAME/my-strategy-project.git
+   git push -u origin main
+   ```
+
+### Connect to Vercel
+1. Go to https://vercel.com/new
+2. Click "Import Git Repository"
+3. Select your strategy project repository
+4. **Important**: Set the **Root Directory** to `dashboard`
+5. Vercel will auto-detect the build settings from `vercel.json`
+6. Click "Deploy"
+7. Wait 1-2 minutes — your dashboard will be live at a `.vercel.app` URL
+
+### Auto-Deploy
+Every time you push changes to GitHub (which `/session-end` does automatically), Vercel will rebuild and update your dashboard. No extra steps needed.
+
+### Privacy Warning
+**Your dashboard will be publicly accessible** at the Vercel URL unless you:
+- Use Vercel's password protection (requires a paid plan)
+- Keep the URL private and don't share it
+- Use a private GitHub repo (the code is private, but the deployed site is still accessible to anyone with the URL)
+
+For sensitive strategy work, consider whether a live dashboard is appropriate for your situation.
+
+---
+
+## Step 8: Start Working
+
+Every time you sit down to work on your project:
+
+1. Open Terminal, navigate to your project folder
+2. Start Claude: `claude`
+3. Tell Claude: `/session-start`
+4. Claude will load your context and show you what's happening
+
+When you're done:
+1. Tell Claude: `/session-end`
+2. Claude will save everything, update the dashboard, and commit your changes
+
+---
+
+## What Each Skill Does
+
+Here's a plain-language guide to every skill available:
+
+### Setup & Session
+| Skill | What it does |
+|-------|-------------|
+| `/onboard` | Sets up a new project from the template (run once) |
+| `/session-start` | Opens your project — loads context, shows priorities, checks for conflicts |
+| `/session-end` | Closes your session — saves everything, updates dashboard, commits changes |
+| `/health-check` | Runs a diagnostic on your project — finds issues and suggests fixes |
+| `/rebuild-snapshots` | Refreshes the quick-loading summaries from your raw files |
+
+### Research & Discovery
+| Skill | What it does |
+|-------|-------------|
+| `/enrich-entity` | Deeply researches a competitor or prospect — fills in their dashboard profile |
+| `/synthesise` | Pulls together research from multiple files into a structured summary |
+| `/process-call` | After a meeting or call, structures your notes and updates everything automatically |
+
+### Pipeline
+| Skill | What it does |
+|-------|-------------|
+| `/pipeline-update` | Updates a prospect's status (e.g., "meeting booked" to "meeting completed") |
+| `/outreach-sequence` | Designs a series of messages to reach out to a prospect |
+
+### Analysis & Decisions
+| Skill | What it does |
+|-------|-------------|
+| `/critical-reasoning` | Pressure-tests an idea — finds blind spots, challenges assumptions |
+| `/decision` | Records a strategic decision with full context so you remember why you chose it |
+| `/compare-options` | Scores and compares 2-5 strategic options side by side |
+
+### Reporting
+| Skill | What it does |
+|-------|-------------|
+| `/weekly-report` | Generates a summary of the week's progress for stakeholders |
+
+---
+
+## Things You Can Say to Claude
+
+You don't need to memorise skills. You can talk to Claude in plain English and it will figure out what to do. Here are some real examples:
+
+**After a meeting:**
+> "Here's my meeting transcript from the call with Acme Corp. Can you analyse this against my existing research to see what it means for our current hypothesis?"
+
+**When researching:**
+> "Pull together everything we know about the competitive landscape and tell me where the biggest white space is."
+
+> "I just found this industry report — read it and update our competitor research with anything new."
+
+**When you need to decide something:**
+> "We need to decide between building our own data pipeline vs. using a third-party vendor. Set up a comparison with the pros, cons, and scores."
+
+> "Play devil's advocate on our go-to-market strategy. What are we not seeing?"
+
+**Before a meeting:**
+> "Prep me for my call with Nexus Payments tomorrow — what do we know about them and what should I ask?"
+
+**Pipeline management:**
+> "I have 3 new leads from a conference. Add them to the pipeline and draft a personalised outreach sequence for each."
+
+> "Based on all the discovery calls so far, what patterns are emerging? Are we hearing the same pain points?"
+
+**Getting a status update:**
+> "Write a 2-paragraph update I can send to the board summarising where we are this week."
+
+> "How close are we to hitting any of our kill conditions?"
+
+The key idea: Claude has your full project context loaded — research, pipeline, decisions, hypothesis, scoring. So you can ask strategic questions and get answers grounded in your actual data, not generic advice.
+
+---
+
+## Common Questions
+
+**Q: Do I need to know how to code?**
+No. You talk to Claude in plain English. The only "commands" you need are the skill names (like `/onboard` or `/session-start`).
+
+**Q: Can I use this for personal projects?**
+Yes. It works for any project where you're making strategic decisions based on research — from choosing a business idea to evaluating job offers.
+
+**Q: What if I make a mistake?**
+Everything is tracked in Git (version control). You can always go back to a previous state. Ask Claude: "Can you undo the last change?"
+
+**Q: Can multiple people work on the same project?**
+Yes. The STATUS.md file coordinates who's working on what. Each person runs Claude Code in their own terminal.
+
+**Q: How do I update to the latest template version?**
+Check the CHANGELOG.md in the original repository for what's new. For major updates, it's often easiest to start fresh and re-run `/onboard`.
+
+**Q: Is my data private?**
+Your project files live on your computer and (optionally) in your private GitHub repository. If you deploy a dashboard to Vercel, the dashboard URL is accessible to anyone who has it. Don't put confidential information in the dashboard if you're concerned about privacy.

package/template/docs/memos/evidence-grading.md

@@ -0,0 +1,27 @@
+# Evidence Grading & Contradiction Resolution
+
+## Evidence Tags
+
+Every factual claim in memory files, the executive summary, and memos must be tagged:
+
+- **[CONFIRMED]** — Validated by 2+ independent sources or primary data (discovery call)
+- **[SECONDARY]** — From one credible source (industry report, vendor docs, public financials)
+- **[INFERENCE]** — Logical deduction from confirmed/secondary evidence, not directly observed
+- **[ASSUMPTION]** — Untested belief. Must be flagged for validation in discovery.
+- **[STALE]** — Was confirmed but may have changed. Needs re-verification.
+
+## Grading Rules
+
+- When synthesising across files, weight CONFIRMED > SECONDARY > INFERENCE > ASSUMPTION
+- Never elevate an INFERENCE to CONFIRMED without new evidence
+- When two sources conflict, flag the conflict explicitly — do not silently pick one
+- Discovery call data is always CONFIRMED (primary). Web search results are SECONDARY.
+
+## Contradiction Resolution
+
+When memory files or research produce findings that conflict with the executive summary:
+
+1. Flag the contradiction explicitly — do not silently ignore it
+2. The executive summary is NOT automatically correct. It reflects the last hypothesis review.
+3. Contradictions are resolved in the NEXT hypothesis review, not by editing files ad hoc.
+4. Until resolved, both positions must be visible. Add a "Pending reconciliation" note.

package/template/docs/memos/housekeeping-reference.md

@@ -0,0 +1,101 @@
+# End-of-Session Housekeeping — Full Reference
+
+Every session that produces or modifies files MUST run this checklist before finishing.
+
+---
+
+## Step 1: File Placement Audit
+
+Scan all files created or modified this session. Verify each one:
+
+| File type | Correct location | Template used? |
+|-----------|-----------------|----------------|
+| Entity teardown | `research/competitors/<slug>.md` | `templates/entity-teardown.md` |
+| Market/TAM research | `research/market/` | — |
+| Technical research | `research/technical/` | — |
+| Call notes | `discovery/calls/<entity>-<date>.md` | `templates/call-notes.md` |
+| Call prep | `discovery/prep/<entity>-prep.md` | `templates/call-prep.md` |
+| Strategy memo | `docs/memos/` | — |
+| Deck/presentation | `docs/decks/` | — |
+| Data file (CSV/Excel) | `data/` | — |
+
+If a file is in the wrong place, move it now. If created without the correct template, flag it.
+
+## Step 2: Memory Cascade Update
+
+| Work done | Update required |
+|-----------|----------------|
+| Competitor/entity research | Add/update row in `memory/research.md` |
+| New target {{ENTITY_TYPE_PLURAL}} found | Update Target Entity List in `memory/research.md` |
+| Technical pathway research | Update Technical Findings in `memory/research.md` |
+| Outreach sent (any channel) | Update Outreach Log in `memory/discovery.md` + update `memory/MEMORY.md` Discovery line |
+| Call prep created | Note in pipeline tracker for the matching {{ENTITY_TYPE}} entry |
+| Call notes processed | Note in pipeline tracker for the matching {{ENTITY_TYPE}} entry |
+| {{ENTITY_TYPE}} disqualified | Update `memory/MEMORY.md` Discovery line + note reason in `memory/discovery.md` Outreach Log |
+| Customer discovery call | Update Pain Point Rankings + WTP Signals in `memory/discovery.md`. Update `memory/MEMORY.md` Discovery line. |
+| Kill condition evidence changed | Update Kill Condition Tracker in `memory/discovery.md` + update `memory/MEMORY.md` Kill Conditions table |
+| Strategic decision made | Add entry to `memory/decisions.md` + one-liner to `memory/MEMORY.md` |
+| Option scoring changed | Update `memory/scoring.md` |
+
+## Step 3: Hypothesis Review & Executive Summary Update
+
+Ask: "Did this session produce new evidence that could affect hypothesis confidence?"
+
+- If **YES** → run `/critical-reasoning` before updating exec summary. Only update after user confirms the briefing.
+- If **NO** (structural/setup work) → skip to Step 4.
+
+## Step 4: Status Blurb Update
+
+Update `docs/output/status-blurb.md` every session. Rules:
+- Max 4 sentences
+- Plain English — no jargon (no technical codes, no internal terminology)
+- Never mention internal tooling, agent workflows, file reorganisation, or process changes
+- Cover: current phase, what's active, key unknown, any blockers
+
+## Step 4b: README.md Status Sync
+
+Copy the content of `docs/output/status-blurb.md` into the `<!-- STATUS:START -->` / `<!-- STATUS:END -->` block in `README.md`. This keeps the GitHub landing page current.
+
+## Step 5: Project Activity Update
+
+Assess whether this session produced exec-worthy progress. Apply the filter:
+- **Log**: Concrete deliverables, milestones reached, external-facing actions (outreach sent, calls completed, research synthesis, memos)
+- **Skip**: Dashboard UI work, CSS fixes, skill creation/updates, data schema changes, call prep creation, internal tooling, maintenance, housekeeping
+- **Rule of thumb**: If you wouldn't mention it in a 2-minute standup with an investor, don't log it
+
+## Step 6: STATUS.md & Work Log Update
+
+Every session that completes work items MUST do ALL of the following:
+
+a) Move completed "In Progress" rows → STATUS.md "Completed Recently"
+b) Append the SAME rows to `docs/output/work-log.md` (append-only, never trim)
+c) Trim STATUS.md "Completed Recently" entries older than 1 week
+   (safe to trim — they're permanently in work-log.md)
+d) Remove "Blocked / Waiting" entries that are resolved
+
+Steps (a) and (b) happen together — never do one without the other.
+
+## Step 7: Rebuild Dashboard Data
+
+Always run `cd dashboard && node build-data.js` before committing. This regenerates all dashboard JSON from source-of-truth files. Include regenerated `dashboard/data/*.json` files in the commit.
+
+## Step 8: Structure Health Check
+
+Run only if 3+ new files created or project has grown significantly:
+- Files in root that belong in a subfolder?
+- Files in wrong subfolder?
+- Duplicates that should be merged?
+- **Versioned files (v2, v3, _old, _new, _backup)?** Keep only latest, rename clean, delete outdated.
+- Templates need updating?
+- `memory/MEMORY.md` still under 200 lines?
+
+## Step 9: Commit & Push
+
+After all updates are complete:
+1. Stage all files modified during the session and housekeeping, **including regenerated `dashboard/data/*.json` files** (use specific file paths, never `git add -A`)
+2. Commit with a concise message summarising the session's work (not the housekeeping itself)
+3. Push to remote
+
+## Step 10: Report
+
+Report housekeeping outcome to user.

package/template/docs/memos/reference-context.md

@@ -0,0 +1,30 @@
+# Reference Context — {{PROJECT_NAME}}
+
+Static facts extracted to reduce context loading. Update when underlying data changes.
+
+---
+
+## Key Metrics
+
+<!-- Market numbers that any session might need. Keep to 10–15 facts. -->
+
+| Metric | Value | Source | Evidence Grade |
+|--------|-------|--------|---------------|
+
+---
+
+## Competitive Landscape (Summary)
+
+<!-- One-liner per competitor. Full teardowns in research/competitors/ -->
+
+| Competitor | What they do | Why they matter | Gap we exploit |
+|-----------|-------------|----------------|---------------|
+
+---
+
+## Scenario Planning
+
+<!-- If the strategy depends on external conditions, document scenarios here -->
+
+| Scenario | Probability | Impact on Strategy | Response |
+|----------|------------|-------------------|----------|

package/template/docs/output/project-activity.md

@@ -0,0 +1,8 @@
+# Project Activity — {{PROJECT_NAME}}
+
+Exec-worthy milestones and deliverables only. If you wouldn't mention it in a 2-minute investor standup, don't log it here.
+
+---
+
+| Date | Milestone | Details |
+|------|-----------|---------|

package/template/docs/output/status-blurb.md

@@ -0,0 +1,4 @@
+<!-- Max 4 sentences. Plain English. No jargon. No internal terminology.
+     Cover: current phase, what's active, key unknown, any blockers. -->
+
+Project not yet initialised. Replace placeholders and begin work.

package/template/docs/output/work-log.md

@@ -0,0 +1,8 @@
+# Work Log — {{PROJECT_NAME}}
+
+Append-only permanent log of all completed work. Sourced from STATUS.md "Completed Recently" rows.
+
+---
+
+| Workstream | Item | Completed | Output |
+|-----------|------|-----------|--------|