New Second Brain Update, Open Source Skills, and a Request: Show Us How You Use It
We just published a public GitHub repo with ready-to-use Claude skills for Second Brain.
github.com/HEXACT-INC/second-brain-claude-skills
A skill is a markdown file that teaches Claude how to handle a specific task with your data. Drop it into your skills folder, restart Claude, and it works.
The repo currently have five skills:
second-brain is the core skill. It gives Claude safe, permission-first access to your local database. Read anything, write only when you say so. This one should be installed first. Everything else builds on top of it.
quick-lookup answers “what do I know about X?” questions. Ask about a person, a company, a deal, and it pulls everything from your KB, synthesizes it into a short answer. Useful before calls, before signing contracts, before replying to an email you’re not sure about.
weekly-review generates a summary of everything you captured in the last 7 days. Notes, meetings, saved ideas, new contacts. Instead of scrolling through your database wondering what happened last week, you get a concise brief.
meeting-prep builds a pre-call brief for a specific person or company. It pulls contact history, past meetings, relevant notes, payment data if it exists. You walk into the call knowing what matters.
save-and-categorize handles “save this” requests. Share a thought, a link, a meeting takeaway, and the skill figures out the right entry type, writes a clean title, and stores it properly. No manual tagging.
Tell Us What’s Missing, What’s Blocking You, or What You’d Change
We shipped the app and the skills based on what we already know. But you’re using it differently. You loaded different data, you run a different business, you ask different questions. That means you’re hitting walls we haven’t hit yet.
What feature would make the biggest difference for you right now? What’s confusing? What workflow are you trying to build but can’t? What skill would save you time if it existed? A daily journal, a deal tracker, an invoice reviewer, a hiring pipeline brief, a content calendar. If it involves structured data and a repeatable task, it’s probably a good skill. Reply to this email or leave a comment.
If you've built a skill that works for you and you're willing to share it, the repo is open. Fork it, add your skill folder with a SKILL.md and a README.md, and open a pull request. Same goes for improving the existing ones. One rule: keep each skill focused on one job.
Any feedback that points to a product gap - we’ll build it. That’s how the new import formats, database management tool and global search happened. Users asked, we shipped.
New Release: Download the Latest Version
We also shipped a new version of Second Brain today. If you already have the app installed, download the latest build from brain.hexact.io. Here’s what changed:
Database management tool. You can now edit any table in your database manually, right from the app. No SQL required. This was the most requested feature after launch.
Global search. Search across all your local records from one search bar. Contacts, notes, articles, transactions. Type a name or a keyword and see results from every table.
Bug fixes. Stability improvements across import pipelines, dashboard rendering, and MCP connectivity.
If you’re on an older version, the update is worth it. The database tool alone changes how you interact with your data day to day.
Use Cases That Make This Click
I’ve been using Second Brain daily since before the launch. Here are some real workflows that show what this setup can do once your data is loaded.
Pre-meeting intelligence. Before any call, I ask Claude: “What do I know about this person?” It pulls their contact record, any past emails, payment history if they’re a customer, notes from previous meetings. I walk in prepared without spending 15 minutes digging through apps.
Customer health check. “Show me all customers who paid more than $500 but haven’t logged in since 2024.” That query runs against my actual database. Real data, real answers. No dashboard needed.
Content research from your own writing. “What have I written about data ownership?” Claude searches my 168 published articles and pulls the relevant ones. I see my own arguments, positions, and examples. Useful to see how things change compared to what I said six months ago.
Supplier and vendor comparison. If you use Hexomatic to scrape pricing pages or product catalogs, that data flows into Second Brain automatically. Ask Claude to compare suppliers by price, minimum order, or any other detail. The comparison runs on your scraped data, not on Claude’s general knowledge.
Trend spotting across scraped data. Scrape job postings in your industry weekly. After a few months, ask Claude: “Are companies hiring more for AI roles compared to three months ago?” You get trend analysis from your own data, not a generic report.
The pattern is the same in every case. Your data is already in the database. You ask a question in plain language. Claude reads your actual records and answers. No exports, no spreadsheets, no switching between six apps.
Get Started
Download the latest version at brain.hexact.io
Install the
second-brainskill, then add whichever specialized skills you wantBrowse the full repo at github.com/HEXACT-INC/second-brain-claude-skills
Three things you can do right now: update the app, try a new skill, and tell us what you’re building with it. Reply to this email with your use case, open a pull request with your skill, or just star the repo so we know you’re watching. We’re building this in the open and your input shapes what comes next.
Interesting concept, thank you for sharing. Still, one question: is my data safe, does it stay local? If not, how can I run Second Brain with local LLMs?