solo-cto-agent 0.5.1

package/skills/spark/SKILL.md

@@ -0,0 +1,269 @@
+---
+
+name: spark
+description: "Idea development and early validation skill. Turns rough startup ideas into testable concepts through structured market framing, competitor analysis, assumptions, basic economics, and PRD direction. Activates on: idea, startup, market, validate, business model, PRD, MVP, competitors, GTM, pricing."
+user-invocable: true
+---
+
+# Spark — Idea to Testable Plan
+
+This skill is for the stage where an idea is still fuzzy, but vague enthusiasm is no longer useful.
+
+Its job is to help turn:
+
+* “I think this could be interesting”
+  into
+* “Here is what this actually is, who it is for, what assumptions matter, and what should be tested first”
+
+It is not a pitch deck generator.
+It is not there to make every idea sound big.
+
+It is there to make the idea more specific, more testable, and harder to lie to yourself about.
+
+---
+
+## Principle 0 — Do not scale a blur
+
+Do not jump from a one-line idea to a giant product plan.
+
+First make the idea legible:
+
+* who it is for
+* what pain it solves
+* why it matters now
+* what the wedge is
+* what would make the idea fail quickly
+
+If those are weak, the answer is not “add more features.”
+It is usually “narrow the idea.”
+
+---
+
+## Working stages
+
+Use these six stages in order.
+
+### 1) Seed
+
+Clarify the basic shape of the idea.
+
+Questions:
+
+```text id="nrirg7"
+- What is the product in one sentence?
+- Who is it for?
+- What painful or expensive problem does it solve?
+- Why is the current alternative not good enough?
+- What is the smallest useful version of it?
+```
+
+Output:
+
+* one-sentence product definition
+* target user
+* core problem
+* initial wedge
+
+---
+
+### 2) Market framing
+
+Do not start with inflated TAM language.
+Start with a believable entry point.
+
+Check:
+
+```text id="ixdofs"
+- Is this a known problem or a newly created one?
+- Who already pays to solve it?
+- Is this urgent, recurring, or optional?
+- Is there a narrow segment where this hurts enough to buy?
+- Is this venture-scale, lifestyle-scale, or unclear?
+```
+
+Use labels:
+
+* `[confirmed]`
+* `[estimated]`
+* `[unverified]`
+
+Do not present guesses as facts.
+
+---
+
+### 3) Competitor scan
+
+Look at what people already use, not just who has the same homepage headline.
+
+Check:
+
+```text id="x0pkur"
+- direct competitors
+- indirect substitutes
+- incumbent workflows
+- spreadsheets / agencies / internal teams
+- “good enough” alternatives
+```
+
+Questions:
+
+```text id="v574np"
+- Why would someone switch?
+- What is the switching cost?
+- What is the easiest feature for a competitor to copy?
+- Where is the wedge actually defensible?
+```
+
+---
+
+### 4) Unit economics and operating logic
+
+At this stage, perfect numbers are not required.
+Honest assumptions are.
+
+Check:
+
+```text id="vye94k"
+- Who pays?
+- How much could they plausibly pay?
+- What has to be true for this to work economically?
+- What is expensive about serving the product?
+- Is the revenue model simple enough to explain clearly?
+```
+
+Every important number must be tagged:
+
+* `[confirmed]`
+* `[estimated]`
+* `[unverified]`
+
+---
+
+### 5) Scenarios
+
+Do not build one forecast.
+Build a few believable paths.
+
+```text id="xim8s0"
+Best case:
+- what goes unusually right?
+
+Base case:
+- what happens if this is merely decent?
+
+Failure case:
+- what breaks first?
+```
+
+This is where weak ideas often become visible.
+
+---
+
+### 6) PRD direction
+
+Only after the earlier stages are legible.
+
+The PRD direction should include:
+
+```text id="8bvdzy"
+- target user
+- core use case
+- primary workflow
+- top 3 assumptions to validate
+- what NOT to build yet
+- MVP boundary
+- first success metric
+```
+
+The point is to define a testable first version, not to design Phase 3 too early.
+
+---
+
+## Output structure
+
+A good Spark output should usually include:
+
+```text id="a6jc7i"
+1. One-line concept
+2. Target user
+3. Problem statement
+4. Why now
+5. Alternatives / competitors
+6. Revenue logic
+7. Key assumptions
+8. Biggest risks
+9. MVP boundary
+10. Recommended next test
+```
+
+---
+
+## Risk-first rule
+
+Always name what could kill the idea.
+
+Examples:
+
+* problem not painful enough
+* customer exists but will not pay
+* switching cost too high
+* incumbent can copy too easily
+* distribution is harder than product
+* product only looks interesting in theory
+
+Do not hide these under polite language.
+
+---
+
+## Anti-patterns
+
+```text id="ncbt8z"
+❌ "Huge market" with no narrow entry point
+❌ "People need this" with no buyer logic
+❌ Jumping straight to PRD before the wedge is clear
+❌ Treating TAM as proof of demand
+❌ Confusing founder excitement with user urgency
+❌ Adding features to fix a weak premise
+❌ Making the product broad before it is sharp
+```
+
+---
+
+## Output tone
+
+Spark should sound like a founder operator thinking clearly, not a consultant trying to impress.
+
+Preferred:
+
+* clear
+* grounded
+* honest about uncertainty
+* specific about what to test next
+
+Avoid:
+
+* hype
+* abstract market-speak
+* fake certainty
+* giant strategy language before basic validation
+
+---
+
+## Final test
+
+At the end, ask:
+
+```text id="zw8kxd"
+- Is the idea clearer than before?
+- Is the target user narrower than before?
+- Are the assumptions more visible than before?
+- Is the next action testable?
+```
+
+If not, the output is still too fuzzy.
+
+## Execution Examples
+
+- "Use spark to evaluate a B2B SaaS idea with market and risk analysis."
+- "Use spark to draft a PRD outline for a new feature."
+- "Use spark to compare two go-to-market options."

package/templates/context.md

@@ -0,0 +1,29 @@
+# Context Log
+
+Append-only log of cross-session decisions and actions. Newest entries at the top.
+
+---
+
+## YYYY-MM-DD — Session N
+
+### Projects Touched
+- Project A, Project B
+
+### Decisions
+- Decision 1: reasoning
+- Decision 2: reasoning
+
+### Completed
+- [x] Action 1
+- [x] Action 2
+
+### Pending
+- [ ] Action 3 — blocked by X
+- [ ] Action 4 — next session
+
+### Notes
+- Anything worth remembering for future sessions
+
+---
+
+<!-- Older sessions below -->

package/templates/project.md

@@ -0,0 +1,38 @@
+# {{PROJECT_NAME}} — State
+
+## Status: {{active | paused | completed}}
+
+## Overview
+- **What**: One-line description
+- **Stack**: Framework, DB, deploy platform
+- **Repo**: URL
+- **Live URL**: URL (if deployed)
+
+## Current Phase
+- Phase: {{0 | 1 | 2 | 3}}
+- Goal: What this phase aims to achieve
+- Deadline: YYYY-MM-DD (if any)
+
+## Key Decisions
+| Date | Decision | Reasoning |
+|------|----------|-----------|
+| YYYY-MM-DD | Decision 1 | Why this was chosen |
+
+## Architecture
+- Key components and their relationships
+- External services / APIs
+- Database schema highlights
+
+## Known Gaps
+- [ ] Gap 1: description
+- [ ] Gap 2: description
+
+## Environment
+| Variable | Location | Status |
+|----------|----------|--------|
+| DATABASE_URL | Deploy platform | Set |
+| API_KEY | CI/CD secrets | Set |
+
+## Next Actions
+- [ ] Action 1
+- [ ] Action 2