Hi! I am Ricardo Sueiras, leading SDLC transformation at Indicium AI. That means I get to work with customers and help them understand how to use agentic AI to build software. New tools, practices and opinions turn up every week, and Build without Limits cuts through the noise. In each edition we put together a short, curated briefing on what's worth reading, watching and worth trying. You can skim it in five minutes and dive deeper wherever something catches your eye.
Top of Mind
Top of mind this week is Developer Portals (sometimes referred to as IdP's, although as someone who has spent time building identity systems, I always get the hump that this TLA has been usurped!). I had several good discussions with customers, who are looking at leveraging their developer portals to include gold paths for agentic solutions.
One of the surprising things that has come up in these conversations is how to rethink those same developer portals for a new kind of developer. Citizen Developer specifically, and how platform teams are now beginning to rethink what the design and user experience looks like for this new kind of user. Hannah Foxwell added to this discussion in her post on LinkedIn by adding that it was not just Citizen developer we needed to think of but agents more broadly. This makes sense to me, one place in which to manage and govern everything. I am folllowing this closely, and will share what I find.
Must reads this week
These were my favorite posts this week.
Amazon rewrites its software engineering playbook for an era when AI writes the code - took me by surprise last week, with Marc Brooker sharing some of the lessons Amazon has learned across their engineering teams and how this is changing how they approach development.
https://github.com/adimango/ai-adoption-playbook - provides an opinionated guide on how you might roll out AI across your organisation. I haven't tried it yet, and I have my reservations on how this might work as in my experience, how you roll out and enable an organisation for AI varies hugely and is based on lots of factors. That said, they have done a lot of work pulling together useful tools that will help you with the planning activities.
Recap from GitHub universe- if you want a summary of the main things that went down at Github's annual tech conference, I found this post provided a good summary. I also listened to the Syntax podcast as well. There have been lots of stories about Github's future as it tries to cope with the sheer scale of the impact of agentic engineering. Inside GitHub’s new Git storage architecture looks at how they are innovating their storage architecture to cope.
Scaling software built with AI: where the real bottlenecks are - a long read that covers a lot of ground, looking at the current state of agentic development. Whilst there is nothing new or groundbreaking in this post, its ways its all brought together makes it a good read.
Watch this
Last week I listened to this episode of the Engineering Enablement podcast from the folks at DX, where Roblox's Andrew Swerdlow dived into how they approach code review across their engineering teams and how they measure the impact of agentic engineering. This one is packed with some great approaches that I will be adding to my toolbox.
Tools to try out
Third-party tools links are provided as-is. They belong to their owners, follow their own licenses, and are used at your own risk.
https://github.com/shanselman/sandbox-mcp - this was news to me, that Windows had a capability for creating sandbox environments that you can use to launch your agents in. None other than Microsofts Scott Hanselman has put together an MCP that gives your agent everything you need to get started. Many of the developers I work with use Windows, so this is a great tip.
https://github.com/govijr/terminal-mac - if you like the Windows sandbox tool above, then this tool provides a separate desktop for your computer use agents. The guest is a separate Mac: it has its own apps, files, accounts and permissions. Nice detailed docs, and this tool already has 2.5K stars.
https://github.com/richardedens/crissp - is another AI SLDC framework that has three layers (agent colleagues that build, guards that make sure it is built correctly , and then product which defines what gets built). Not had time to try it out, but I like to check the repos out and get inspired or borrow ideas from them.
From the field
Developer experience is increasingly surfacing in the work I do with customers, which should perhaps be no surprise for folk who are working at helping organisations adopt agentic engineering. Recently it has come up in the topic of quality, and how a good developer experience correlates to higher quality software. We knew this of course, and Accelerate and DORA provided the data to back this up, but as you start to think about your own agentic engineering roll outs and SDLC modernisation, make sure you put developer experience at the centre.
One of the ways I see developer experience manifest at a practical level is how they are setup for agentic development. When I speak with developers, I am pretty nosey and like to see their setup (and I also find that developers mostley love to share). When I saw this thread come up on Hacker News of developers sharing their setup, I spent a lot of time and went down lots of rabbit holes to find out just how developers were optimising for agentic development. When I last checked it, just over 600 responses.
If you enjoyed this, then remember to subscribe to the IndiciumAI blog so that you can keep up to date with future editions of this newsletter, as well as other great content that technologists at IndiciumAI will publish.
Top comments (0)