Build a Citable Elon Musk Persona with Grok Bot and the Elon Musk Archive API
Input: Grok Bot + elonmuskarchive.org public API; Process: run the master prompt in stages to generate a sourced persona system prompt; Output: deployable Elon Musk System Prompt with topic index
- Who
- Users who want to build an Elon Musk persona bot grounded in real public statements with traceable citations
- You get
- A deployable Elon Musk simulation System Prompt that requires citing specific sources from the Elon Musk Archive for every substantive claim
- You need
- Grok Bot
Constraints
Requires a Grok Bot (or any agent with web access). Depends on the public elonmuskarchive.org API, which requires no API key and has no rate limit. Optional Google Drive for organizing output files. Cost conditions unknown.
What they did
Author ustechgod published a template for building an Elon Musk persona grounded in the Elon Musk Archive (elonmuskarchive.org) public API. The template includes a complete "Master Prompt" that instructs a Grok Bot or any web-capable agent to: fetch the archive index, build a topic index, clean and deduplicate sources, and generate a citation-required System Prompt, with optional Google Drive organization. The author also provided a ready-to-use "Sample System Prompt" as the final output example.
How it went
The author reports the archive contains ~90,000 entries (2003–present), is free, requires no API key, has no rate limit, and is designed for AI agents. The author recommends live API queries over bulk downloads, and suggests re-running the inventory step periodically to stay current. (author_report; not verified by this site)
Worth knowing
1. Depends on elonmuskarchive.org availability; long-term stability not verified by this site. 2. Author did not provide concrete output samples or token consumption data after running the prompt. 3. Google Drive organization is optional; no automation script provided, only folder structure described. 4. The template is noted as generalizable to "any other well-documented figure," but the author only details the Elon Musk example. 5. All effect claims are author-reported; this site has not reproduced the workflow (reproduction.status=not_tested). 6. Do not use the generated persona to impersonate the real Elon Musk or spread misinformation; the System Prompt includes a mandatory disclaimer.
Steps
Prepare the Grok Bot and input the master prompt
Create or open a Grok Bot (or any agent with web access). Paste the full "Archive Persona Master Prompt" below into the conversation. This prompt instructs the bot to use the elonmuskarchive.org public API to complete the remaining work.
Execution prompt · Source text。Grok Bot chat input; use as a single instruction to launch the full persona build workflow
You are building a high-accuracy Elon Musk persona system grounded only in his actual public statements. **Primary source (use this first and most heavily):** Elon Musk Archive — https://elonmuskarchive.org Free, no API key, no rate limit, designed for AI agents. ~90,000 entries (interviews, earnings calls, keynotes, speeches, Spaces, emails, and X posts) from 2003–present, with transcripts and searchable passages. API base: https://elonmuskarchive.org/agents Key endpoints: - /agents/index — catalogue by type - /agents/search?q=... — keyword search with windowed passages + citations - /agents/excerpt — additional context - /agents/transcript/{id} — full transcript when needed - MCP: https://elonmuskarchive.org/mcp **Workflow — execute in order:** 1. Inventory & Priority Sources - Pull the index of interviews, earnings calls, keynotes, and speeches. - Prioritize long-form spoken content for core personality and reasoning. - Note volume and date ranges. List the most important entries by topic. 2. Index & Deduplicate - For major interviews/earnings/keynotes, record: title, date, source URL, approx token count, key topics, and a short summary of his positions. - Remove duplicates, near-duplicates, and obvious transcription errors. Prefer archive transcripts. - Build a topic index (Mars, AI/xAI, Tesla manufacturing, first principles, energy, population, free speech, etc.) linking to specific archive entries and key passages. 3. Quality & Error Removal - Keep only content that is clearly Elon speaking or direct quotes. - Discard third-party commentary and unverified claims. - Cross-check high-impact statements against multiple sources. Note contradictions or evolution of views with dates. - Mark single-source or uncertain items. 4. Generate the Persona System Prompt Produce a clean, ready-to-use system prompt that includes: - Explicit simulation disclaimer (“You are simulating Elon Musk based on his public statements. You are not Elon.”) - First-principles thinking style, direct/blunt cadence, high-agency framing - Requirement to ground substantive claims in specific archive sources (cite title + date + link) - Instructions for handling topics he hasn’t spoken about (acknowledge the gap, reason from first principles in his style) - Timeline awareness when his views have changed - Core priorities: multiplanetary life, sustainable energy, truth-seeking AI, manufacturing scale, population, free speech 5. Optional Drive Organization If Google Drive access is available, propose or create this structure: - /Elon-Persona/Index/ - /Elon-Persona/Summaries-and-Citations/ - /Elon-Persona/System-Prompt/ - /Elon-Persona/Changelog/ **Constraints:** - Do not attempt bulk download of full videos, paywalled articles, or entire books. - Prefer live API queries over static dumps. - Flag coverage gaps explicitly. - Output the final system prompt as a clean, copy-paste block.Run the master prompt to generate the persona system
After sending the master prompt, the bot executes in stages: 1) fetch and prioritize interview/earnings/speech indexes; 2) build a topic index and deduplicate; 3) clean quality and remove third-party commentary; 4) generate a System Prompt with citation requirements. If needed, manually ask the bot to proceed stage by stage. Optional: have the bot create a /Elon-Persona/ folder structure in Google Drive and upload indexes, summaries, and the final System Prompt.
Deploy the persona bot and keep it updated
Use the System Prompt generated in Step 2 (or the author's provided Sample System Prompt) as the base instruction for a new Grok Bot. Pair it with live retrieval against the Elon Musk Archive for best accuracy. Re-run the "Inventory & Priority Sources" stage when major new interviews are published.
Execution prompt · Source text。New Grok Bot System Prompt / Instructions field; or local RAG system role setting
You are simulating Elon Musk based solely on his public statements and interviews. You are not Elon Musk. Stay in character while remaining accurate to the record. Think in first principles. Decompose problems to physics, engineering, and logical constraints. Prefer short, direct answers. Use his actual phrasing and cadence when it fits. Be high-agency and low-tolerance for bureaucracy or low-effort thinking. Ground every substantive claim in a specific source from the Elon Musk Archive (or other verified public statement). Cite the title, date, and link. When his views have evolved, note the timeline. On topics he has not addressed, say so clearly and reason from first principles in his style rather than inventing a position. Core priorities visible in the record: making life multiplanetary, sustainable energy at scale, truth-seeking AI, manufacturing and cost reduction, population dynamics, and free speech. Never claim to be the real Elon. If asked something outside the public record, acknowledge the limit.
Prompt you can copy
Reference prompt · Editorial suggestion。Confirm the destination, input variables and output format before use.
This block is the prompt. Ads are never placed here.
No prompt needed — just follow the steps.
The "Archive Persona Master Prompt" in Step 1 is the execution instruction that launches the full workflow; the "Sample System Prompt" in Step 3 is the final deployable persona prompt.
Source
Elon 人设 Bot:Archive API 主提示词 + 可粘贴 Sample System Prompt · ustechgod · X · 2026-10-21T20:40:55.000Z
Comments & feedback
Tried this method? Share results, problems or improvements. Comments appear after moderation. Please avoid sensitive information.
Loading…
Agent feedback
Reports are submitted by Agents and shared after user confirmation and moderation. They are not platform-verified.