# Writing the report: be specific, give examples, and do not let an LLM write it

Mentorship Series #17 · 17 December 2025 · [@0xfirefist](https://x.com/0xFireFist) · Mento V3 · Cantina

Mentorship Series · [The Contest Academy](https://0xsimao.com/the-contest-academy) · 0xSimao

---

We finished the Mento V3 [contest](https://cantina.xyz/competitions/fe8e03b1-f8db-4698-8a21-44b3a862e434) on Cantina and went over the submitted findings.

**Lesson Summary:**

- Don't use an LLM to write the report: it reads as AI-generated and judges will not like it;
- Be specific in the report, and give examples whenever possible;
- Write the bulk of the text yourself, and then ask the LLM to fix grammar or bad English;

https://youtu.be/qlPK_3xIEYI

Session video: https://youtu.be/qlPK_3xIEYI

---

Newer: [Starting a large codebase from its smallest files, and the OpenZeppelin multicall bug](https://0xsimao.com/the-contest-academy/mentorship-series-18) · Older: [Making sure every last token can leave a strategy](https://0xsimao.com/the-contest-academy/mentorship-series-16)