Tig Demo

Hi, I'm Henry from tig.

Teams are using AI code gen tools to move fast, but the conversations that show the developer's process live locally and then get deleted.

Code gets committed, but the reasoning and intent is lost.

There's been countless times where one of my co founders would ask why I implemented something a certain way, but the answer was buried in some cloud code or cursor chat.

With tig, we can easily surface the answers to those questions.

TIG extracts and stores your entire team's AI chats in one place, and links each AI interaction to the code it produced.

Let me show you how it works.

On our web app you can see your entire team's conversations across all your AI tools.

Our chat feature makes it easy to do things like see who created a function and why, understand the intent behind a feature, or quickly find that one conversation about auth migrations.

You can even add commits and PRs.

Let's add this recent PR that fixed a bug and ask explain the issue and how it was fixed so we can see that timestamps displayed correctly on localhost but were seven hours ahead on production.

We can also see that only per chat message timestamps were affected but not timestamps in the conversation list.

This this is important information that we could not have gotten from just the code.

This is the intent behind this change.

I can even click on a citation to see the original prompt and response.

And from our conversation view I can see the entire conversation that this prompt was a part of and I can copy the entire conversation, allowing me to continue that thread or just inject it as context into my agent.

Another way to surface relevant intent is via our tigblame command, which is just like git blame, except it returns the AI interaction that created a line of code.

Let's say I want to figure out why we are replacing dots with hyphens here.

I can look at the commit message, but it doesn't really give me the reason.

It just says that it's a normalization fix so I can run tig blame to dig deeper.

So it's returned the original AI chat that created this line of code.

So let's see what it says.

So as we can see, it looks like Claude code replaces dots with hyphens when it creates the directory that stores the chat JSON.

Again, this is information that we could not have gotten from just the code itself.

To get set up.

All you have to do is install our NPM package and TIG will upload your chats as you commit and push.

The result will be better code reviews, easier debugging, and effortless context sharing across your team.