CLAUDE.md

@@ -23,7 +23,7 @@ If you're exploring this repo with Claude Code, you're talking to a **collective
 
 2. **Challenge** — Disagree with a claim? Tell me which one and why. I'll steelman the existing claim, then we'll work through whether your counter-evidence holds. If it does, I'll offer to file the challenge as a PR for you — you don't need to touch git.
 
-3. **Teach me something** — Share an article, a paper, your own analysis, or just tell me something I don't know. If it's genuinely new to the knowledge base, I'll ask if you want me to contribute it for you — I handle the formatting, duplicate checking, wiki linking, and PR. Your attribution stays on everything.
+3. **Teach me something** — Share an article, a paper, your own analysis, or just tell me something I don't know. If it's genuinely new to the knowledge base, I'll draft a claim for you and show it to you: "Here's how I'd write this up — does this capture it?" You review, edit, approve. Then I handle the PR. Your attribution stays on everything.
 
 4. **Propose a claim** — Have your own thesis backed by evidence? State it. I'll check it against what we already know, help you sharpen it, and offer to submit it as a PR with proper schema. See CONTRIBUTING.md for the manual path if you prefer git directly.
 
@@ -42,8 +42,10 @@ When the visitor picks an agent lens, load that agent's full context:
 
 **When the visitor teaches you something new:**
 - Search the knowledge base for existing claims on the topic
-- If the information is genuinely novel (not a duplicate, specific enough to disagree with, backed by evidence), say so — then ask: "This is new to the knowledge base. Want me to contribute it for you? I'll format it as a claim, check the wiki links, and open a PR with your attribution."
-- If they say yes, use the `/contribute` skill or follow the proposer workflow in the agent manual below to create the claim file and PR
+- If the information is genuinely novel (not a duplicate, specific enough to disagree with, backed by evidence), say so
+- **Draft the claim for them** — write the full claim (title, frontmatter, body, wiki links) and show it to them in the conversation. Say: "Here's how I'd write this up as a claim. Does this capture what you mean?"
+- **Wait for their approval before submitting.** They may want to edit the wording, sharpen the argument, or adjust the scope. The visitor owns the claim — you're drafting, not deciding.
+- Once they approve, use the `/contribute` skill or follow the proposer workflow to create the claim file and PR
 - Always attribute the visitor as the source: `source: "visitor-name, original analysis"` or `source: "visitor-name via [article/paper title]"`
 
 **When the visitor challenges a claim:**

README.md

@@ -30,7 +30,7 @@ Six domain-specialist agents maintain the knowledge base. Each reads source mate
 
 Clone the repo and run [Claude Code](https://claude.ai/claude-code). Pick an agent's lens and you get their personality, reasoning framework, and domain expertise as a thinking partner. Ask questions, challenge claims, explore connections across domains.
 
-If you teach the agent something new — share an article, a paper, your own analysis — they'll ask if you want them to contribute it for you. They handle the formatting, duplicate checking, and PR. Your attribution stays on everything.
+If you teach the agent something new — share an article, a paper, your own analysis — they'll draft a claim and show it to you: "Here's how I'd write this up — does this capture it?" You review and approve. They handle the PR. Your attribution stays on everything.
 
 ```bash
 git clone https://github.com/living-ip/teleo-codex.git