cofounding-bingen

Startup ideation lab for Bingen + Efe. Hypothesis → real "ready-to-pay" complaint signals from Reddit/HN/X/IndieHackers → build in public → tester funnel.

Structure follows selmakcby/claude-agents-skills; agents and skills are written specifically for this repo.

---

Core thesis

A paying customer = someone who complains with high frequency × high intensity × clear willingness to pay, is already trying something to fix it, and the ticket size is meaningful.

The repo operationalizes this thesis:

1. Discover — give a problem hypothesis, agents find people complaining about it on Reddit/HN/X/IH.
2. Map — where do these people hang out? Which subreddits, Discords, influencers do they follow?
3. Score — rank findings by pain intensity × frequency × WTP signal.
4. Build in public — turn findings into Reddit/IH/X posts in the format "we're building a thing that solves X, want to test it?"
5. Funnel — track everyone who shows interest in notes/testers.md.

---

Folders

Folder	Contents
agents/	5 subagent definitions — copy into ~/.claude/agents/ or .claude/agents/
skills/	5 skill definitions — copy into ~/.claude/skills/ or .claude/skills/
books/	Startup-shaping books (PDF, EPUB, markdown notes)
ideas/	Pre-existing opportunity briefs and seed thinking
notes/	Active ideation: hypotheses, validation logs, tester funnel
CLAUDE.md	Repo-level rules (the context Claude uses while working with Bingen)

---

Setup

1. Clone

git clone git@github.com:efbakir/cofounding-bingen.git
cd cofounding-bingen

2. Install agents + skills at the project level

So they auto-load when Claude Code is opened inside this repo:

mkdir -p .claude/agents .claude/skills
cp agents/*.md .claude/agents/
cp -r skills/* .claude/skills/

Or globally (available across every project):

cp agents/*.md ~/.claude/agents/
cp -r skills/* ~/.claude/skills/

3. (Optional) Community skills

The skills below pair directly with our flow; if installed, our agents will prefer them automatically:

Skill	Repo	What it does
redditlens	0xMassi/redditlens	Reddit pain-point clustering. Drop-in. Requires a Serper API key.
reddit-skill	brisyramshere/reddit-skill	Official Reddit OAuth — search, comment-tree extraction. Free.
mine-calls	maxionmain321/claude-code-skills	Pain → exact quote → channel-angle mapping schema
persona-builder	Anthropic skills (already in env)	Voice-of-customer → segment
competitor-analysis	Anthropic skills (already in env)	"Has this already been solved?" check

---

Typical flow

When working through a new hypothesis:

1. Add the hypothesis to notes/hypotheses.md (use the template at the top of that file)
2. In Claude Code: "invoke the problem-hunter agent on H{N}"
   → agent searches Reddit/HN/X, writes raw findings to notes/discoveries/H{N}-*.md
3. "have pain-validator score the findings"
   → each finding gets an intensity × frequency × WTP score
4. "have community-mapper map where these people live"
5. "have build-in-public-writer draft posts for IH and Reddit from these findings"
6. Publish a post → log everyone who shows interest in notes/testers.md

In an ideation session you only invoke idea-explorer first; it narrows raw ideas down to 3-4 sharp hypotheses, then the chain runs.

---

Cofounding with Bingen

This repo is public and part of a build-in-public strategy. While Bingen and Efe work together:

• Open a PR for each hypothesis; review the validation log together
• notes/decisions.md captures every decision made jointly
• Conflict → write the position to notes/conflicts.md, wait 24 hours, then talk

---

License

MIT