Grok BotSecurityVulnerability huntingAI score · 65/100AI-assisted edit · no obvious prompt-injection riskX

Grok Bot Flagged a GitHub Cold-Email Scam Unprompted

Watch how an AI assistant spotted an identity-leasing scam, cancelled a call, and drafted a reply without being asked

BeginnerStepsTip
Who
A developer who received a cold email from GitHub and had already booked a call
You get
Grok Bot auto-flagged the scam, cancelled the call, and replied on the user's behalf
You need
Grok Bot

Constraints

The source only describes the user's observation; it does not explain how Grok Bot was configured. Requires Grok Bot access plus email/calendar integration (mechanism unspecified).

What they did

Developer wyrjerry111 received a cold email from someone claiming to be a Japanese web developer who found them on GitHub and proposed long-term collaboration; a call was already booked for Sunday. Grok Bot analyzed the email without human instruction, detected that the name did not match the email, noted the absence of project-specific references, recognized the pattern as a known identity-leasing scam, cited public cases and government reports, then automatically cancelled the call and drafted a reply.

How it went

Author reports that Grok Bot successfully blocked the suspected scam, cancelled the scheduled call, and composed a reply on their behalf. (author_report)

Worth knowing

Single-user anecdote; not independently verified. Grok Bot's specific trigger mechanism, email access method, and calendar integration are undisclosed. Identity-leasing scams are a known public threat type, some reported cases linked to DPRK IT workers. If attempting similar protection, confirm Grok Bot has the required data access and keep human review for high-risk automated actions.

Steps

  1. First-test confirmation

    Before the first real run, test with sample data and confirm with a human before any send/write/delete to external systems.

  2. Receive the email and note red flags

    Receive a cold email from someone claiming to be a "Japanese web dev" who found you on GitHub and wants to "collaborate long-term." You have already booked a call for Sunday. Watch for these signals: the sender's name does not match the email address; the body mentions none of your actual projects; the phrasing follows a templated "long-term collaboration" invite.

  3. Verify the scam and auto-respond

    Grok Bot flags the risk unprompted: it matches the message against known identity-leasing scam indicators (some cases linked to DPRK IT workers), pulls 10 real cases plus FBI/DOJ reports as supporting evidence, then automatically cancels the scheduled call and drafts a reply email for you.

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.

This case documents product-side autonomous behavior; there is no user-copyable prompt. To build similar protection, you must configure Grok Bot's email access and calendar permissions yourself.

Source

未提示：Grok Bot 识别 GitHub 冷邮件为已知骗局并取消通话代回复 · wyrjerry111 · X · 2026-10-17T03:11:58.000Z

Comments & feedback

Tried this method? Share results, problems or improvements. Comments appear after moderation. Please avoid sensitive information.

Loading…

    Related