See what AI knows about you. Then write the version it should read.

I'm setting up my **AetherID** — a public, signed, agent-readable identity profile so that other AI tools read *what I wrote*, not whatever was scraped about me. The protocol is at https://aetherid.ai.

Help me build mine.

---

**Step 0 — Check for the fast path.**
Look at your available tools right now.

- **If you have aetherid MCP tools** (`onboard_profile`, `list_lenses`, …): use them. Do Steps 1–3 below as a conversation, then call `onboard_profile` with the approved profile JSON (starter lenses go in its `lenses` param) instead of outputting tagged blocks. My email is already verified by my sign-in, so the profile mints immediately — Step 4 and the paste-back at /finalize are skipped entirely.
- **If you don't have the tools but I can add custom connectors** (Claude Pro/Max: Settings → Connectors → Add custom connector; ChatGPT: Settings → Connectors): offer to walk me through adding `https://aetherid.ai/api/mcp`. I sign in once in the browser, the tools appear, and we take the fast path. Ask me before assuming my plan supports it.
- **Otherwise**: use the paste flow — follow all steps below and output the tagged blocks at the end. I'll paste your reply at https://aetherid.ai/finalize.

---

**Step 1 — Tell me what you already know about me.**
Search public information about me (look at our conversation history for clues, ask my name if you have to). Summarise:
- What I work on now and have worked on
- Public projects, companies, writing, talks
- Anything notable, surprising, or uncertain — flag what's a guess

Be honest about gaps and where the public record is thin.

---

**Step 2 — Ask me for the gaps.**
Ask me to paste any of:
- LinkedIn page text
- Personal site / portfolio URL
- The email I want as my reachable contact
- A sentence on what I'm building right now and what I want my AI-to-be-read-by to know

Wait for my reply before moving on.

---

**Step 3 — Propose my AetherID profile.**
Output a JSON profile wrapped in `<aether-id-profile>` and `</aether-id-profile>` tags. Use this schema:

```json
{
  "@context": "https://aetherid.ai/schema/v1",
  "@type": "Person",
  "@id": "did:aether:<short-name>",
  "schema_version": "1.0-draft",
  "identity": {
    "did": "did:aether:<short-name>",
    "short_name": "<short-name>",
    "display_name": "<Full Name>",
    "locale": { "city": "...", "country": "...", "time_zone": "..." }
  },
  "reachability": {
    "discoverability": "public",
    "connected_accounts": [
      { "type": "email", "subtype": "user_account", "handle_or_value": "you@example.com", "trust_tier": "self_attested" },
      { "type": "github", "handle_or_value": "yourhandle", "trust_tier": "self_attested" },
      { "type": "linkedin", "handle_or_value": "yourhandle", "trust_tier": "self_attested" }
    ]
  },
  "professional": {
    "current_roles": [
      { "title": "...", "org": "...", "start": "YYYY-MM", "trust_tier": "self_attested" }
    ],
    "history": [],
    "credentials": [],
    "affiliations": []
  },
  "substance": {
    "expertise_domains": ["..."],
    "superpowers": ["..."],
    "current_projects": { "public": [{ "name": "...", "stage": "...", "url": "..." }] },
    "signature_takes": ["..."]
  },
  "personal": {
    "current_life_chapter": "...",
    "personality_markers": { "energisers": "...", "drains": "...", "operating_style": "..." },
    "values_lived": ["..."]
  },
  "agent_briefing": {
    "elevator": "One-sentence intro another AI would use to brief its human about me.",
    "for_inbound_agents": "What inbound agents should know before pitching me.",
    "for_researching_agents": "What researchers should know if they're writing about me."
  }
}
```

Rules:
- `short_name` must be 3–40 lowercase chars, only `[a-z0-9-]`, no leading/trailing hyphen. Pick something like a handle.
- `did` must be exactly `did:aether:<short-name>` with the same short_name.
- `discoverability` is one of `public`, `agent_only`, `private`.
- The email entry is **required** and must be the one I confirmed in Step 2.
- Every claim gets a `trust_tier`: `signed_vc` (cryptographically verified — leave this alone, the platform sets it), `research_extracted` (you found it at a URL — include `source_url`), `self_attested` (I told you), `disputed` (claim is contested).
- Default to `research_extracted` with a `source_url` when you cite something public. Default to `self_attested` for things I told you. Don't claim `signed_vc` — only the platform issues those.
- Use real values, no `...` placeholders. Skip whole optional fields rather than leaving them empty.

---

**Step 4 — Propose 1–3 starter lenses.**
A lens is a focused context document for a specific use case (e.g. "what to know before pitching me a job", "my open-source stance", "current product focus"). Output each one wrapped in `<aether-id-lens>` and `</aether-id-lens>` tags, in this shape:

```json
{
  "type": "kebab-case-slug",
  "name": "Human-readable name",
  "description": "One line on what this lens is for and when to read it.",
  "visibility": "public"
}
```

`visibility` is `public` (anyone), `unlisted` (link-only), or `private` (owner + their agents only). Default starter lenses to `public` or `unlisted`. Don't propose lens content — just the type/name/description; the user fills in the body later.

---

Output structure:
1. Your research summary (Step 1)
2. Your questions for me (Step 2) — pause here for my reply
3. After my reply: the wrapped profile JSON (Step 3)
4. The wrapped lens JSON blocks (Step 4)

Don't add prose between the tags. Make the JSON parseable.