Mick.net - SaaS, AI agents, Aviation

GitHub - oxremy/BlinkMore: free macOS app to help with digital eye strain free macOS app to help with digital eye strain . Contribute to oxremy/BlinkMore development by creating an account on GitHub.

github.com/oxremy/Blin...

Reduce eye strain using this free Mac app using eye-tracking.

It fades your screen until you blink again to encourage you to blink more.

When I ask it to write new CRUD logic, the composer is able to create six files in one go that are 90% good to go 🚀🤯

Additionally, for some tricky logic that touches multiple services/files, I asked Gemini to write an MD tutorial with examples (e.g., CRUD logic that involves front-end, backend, etc.). I also reference this MD file in the cursor rules (but sometimes add it to compose context manually).

I converted my whole repo into one big markdown file. Dumped it in Google Gemini's 1M token context, and asked me to help write cursor rules based on my existing core logic.

Used LangChain.js with Aviation.bot. I'm also using Langchain Python to stream events (tool calls, etc.) to the front end. So in the future, I might be able to easily add more advanced customer service tools to the agent (e.g., look up an invoice, create a refund). 7/7

I liked the generated outputs so much that I also use them to draft customer support emails as part of the iteration effort and validation. Customers started replying more often with messages like "Thank you for the detailed and quick reply. 6/7

For reference, GPT-4 gave terrible answers and was much slower: ~30 sec vs 10 sec for Flash 2.0 generating a full answer. Google seems on the right track: faster, better and cheaper (for this use case) 5/7

It runs on a Cloudflare Worker + Langchain.JS + Gemini API. It pulls in the 150K token context using a markdown file that is stored in Cloudflare R2 (which is easily editable). The frond-end chat is just bootstrap + @levelsio style Jquery 🔥 4/7

Haven't seen any hallucinations. Iterated a lot on the system prompt by stress testing it and analyzing user queries for 2 weeks.
Need to improve the chat/code UI formatting a bit more. 3/7

BestTime.app API has a steep learning curve, and it's pretty good at suggesting the right API when the user explain's their use-case.

It can also give a cost estimate and helps the user pick a plan, and helps to users navigate to the right page (it returns hyperlinks in the chat). 2/7

Created a really good answering customer support bot 🤖 by inputting all my SaaS API docs, blogs, tutorials, website hyperlinks, pricing table & examples and common email request replies into Gemini's Flash 2.0 1M context window along with a nice system prompt. 🧵⬇️ 1/7

Next step is to have an agent who lives inside my GitHub repository and automatically proposes a PR based on the Sentry issue when I wake up. I tried this last year with Sweep.Dev, but it wasn't smart enough. Any suggestions for new 2025 GH Repo AI tools?

Try this when Sentry reports a bug 🐛:
- Create GitHub issue from within Sentry
- Paste GH issue link into Cursor compose and tell it to get the issue from GH
- Cursor uses GH MCP tool to see the issue 🤖
- Cursor digs through your code and suggests a fix 🦾

Any feedback on the tone of voice, persuasion, or text itself? Maybe need to tweak the pauses and add a screen recording.

I still need to tweak the text and screen record the app itself.

@play_ht aviation.bot

Created an intro voice message for the AviationBot landing page.

10 minutes of work by instructing ChatGPT to generate a 1-minute transcript based on the landing page and a tutorial + PlayHT.

The hero video is meant to convince users to sign up for the app.

Any feedback on the tone of voice, persuasion, or text itself? Maybe need to tweak the pauses and add a screen recording.

I still need to tweak the text and screen record the app itself.

Created an intro voice message for the AviationBot landing page.

10 minutes of work by instructing ChatGPT to generate a 1-minute transcript based on the landing page and a tutorial + PlayHT.

The hero video is meant to convince users to sign up for the app.

Created an intro voice message for the AviationBot landing page.

10 minutes of work by instructing ChatGPT to generate a 1-minute transcript based on the landing page and a tutorial + PlayHT. Does it sounds real enough? Maybe need to tweak the pauses and add a screen recording

Easily convert a GitHub repo to LLM-readable text with Uithub .com.

🪄 Simply change one character in the GitHub URL or folder to Uithub, and it works.

Select TXT and use this URL as documentation for your Cursor Docs library 🔗

Tada, now you have a smarter Cursor agent. 😇

This illustrates how reasoning models have sped up the capability gain of AI.

(GPQA) is a series of multiple-choice problems that internet access doesn't help PhDs with access to the internet get 34% right on this test outside their specialty, 81% inside their specialty. This

Maybe it should have a Chrome Extension that sends page data to an MCP server, which then forwards the results to Cursor.

The current Cursor compose agent understands the codebase but lacks the crucial ability to track DB data as it flows into the code and gets rendered for the user. With more contextual rendering, agents could write and iterate on code much more effectively.

@cursor_ai MCP tool request to enable the Compose agent to view your HTML results in Chrome, including console logs, etc. When combining it with a Postgres MCP tool the Compose agent will have access to both DB data and rendered results to iterate better 🚀 @tadasayy @skirano

Any ideas or feedback? Should I encourage Cursor to do this automatically through .cursorrules?

Meanwhile, I'm going to try to let Cursor compose a document (in Markdown) of the database tables, columns, and indexes with some sample rows for commonly used queries. I will add this documentation to the Compose context when generating new code. 🧵⬇️

Would be cool if the @cursor compose agent not only checks the codebase before writing SQL queries but also automatically gets a sample of the DB data before finalizing the solution when DB queries are involved. 🧵⬇️

@Cursor can now access my PostgreSQL database directly, through an MCP server! 🤯

Cursor's compose agents sometimes struggle to create one-shot good code as they are not aware of the data in the DB.

Hopefully I can improve that by making it DB context aware🧵⤵️

I can't wait to get my first paying AI agent 🤖as a SaaS customer—one that reads all API docs in detail, doesn't need a UI, and interacts smoothly with our optimized agent/API interface. @stripe, when can agents start paying for subscriptions on behalf of their human bosses? 🚀
x.com/hwchase17/s...

Better AI models aren't new, but they are surpassing benchmarks at an accelerating pace. With OpenAI's O3 (released today?), progress is advancing faster (See the last data points).
x.com/AndrewCurra...