I would suggest trying a different harness. I only use OpenCode and haven't run into limits on even the $20 OpenAI plan. I will say that I use it more as an assistant and don't usually just say "hey go do this thing" without a plan broken down into tasks. With that being said, i don't think i have used even 50% of my 5 hour allotment, even when having it do everything.
So all that to say, maybe try other harnesses because it could just be a prompt issue in the harness.
I am also an unabashed opencode shill, so take that into account too.
- MCP - if you need discoverability. a great example for this is something like context7
- CLI/API + Skill: When you have a defined workflow you want to use. For me a good example of this is using the Linear GraphQL API instead of their MCP
I prefer the Linear API because, where I work, we have come up with our way of using the tool so I don't need the entire MCP to do what I need the agent to do. I think for most use cases an MCP is a waste of context.
This is really just boiled down to one thing and it has been the same thing for a very long time. Solve a problem people are willing to pay to have solved. The bigger the problem, the more money you will make. Bigger can mean either more widespread or more valuable.
The hard part is finding a problem and then the actual hard part is coming up with a solution that you can produce for cheaper than someone is willing to pay for it.
Don't get me wrong, this is a cool idea but the main complexity with DAGR is around the SAASM functionality which you can't do on a phone, at least not a publicly available one.
I also agree that you should be able to make a DAGR 2 that is much smaller and more powerful but it would always have to be a standalone device.
Great point — you're absolutely right that SAASM is the critical differentiator, and that's not something any civilian device can replicate. Red Grid MGRS runs on standard L1/L5 GPS, so it doesn't have the anti-spoofing or encrypted PPS capability that makes the DAGR essential in contested environments.
That said, the vast majority of DAGR usage I see isn't in GPS-denied or spoofed environments — it's in land nav training, patrol planning, calling in grids on routine ops, and reporting. For those use cases, a phone with good civilian GPS (3-5m accuracy now with dual-frequency) handles it fine.
I'd love to see a DAGR 2 that's smaller and cheaper while keeping SAASM though...here's to hoping we get more options
One idea where the compute is super front loaded would be to use your compute to generate data sets that you could sell. Burn through the $10k build the data sets, then selling them won't cost much in terms of cloud costs.
I also have zero idea on what type of data someone would be interested in buying but you never know.
Some of these are political so take that into account but these are the ones i listen to:
- Joe Rogan Experience ( a bit of everything)
- Matt & Shane's Secret Podcast (Comedy)
- Called to Communion (Catholic)
- Tucker Carlson (Political)
- Part of the Problem (Political)
- Tom Wood's Show (Political/Cultural)
- Your Welcome with Michael Malice (Political/Cultural)
- The Game with Alex Hormozi (Business)
- History Hyenas (Comedy)
If I am going on a long drive, I will also listen to Hardcore History or Martyr Made. The long episodes help me lose track of time.
So i just made dinner and thought of this question. One thing the LLMs a very good at is "I have X, Y, Z ingredients. what can i make and how can i make it?" It has really helped me start to move more towards buying ingredients as opposed to processed crap.
The whole site can be configured in the astro.config.js file and the rest is markdown.
If you want an example you can look at sst.dev and their github repo at https://github.com/anomalyco/sst/tree/dev/www