Steve Gordon @stevejgordon.bsky.social · 16/09/2026For a while I've been using gpt-5.6-terra to implement OpenSpec specifications pretty successfully. In the last week it's become infuriating. I give it a clear "definition of done - ALL tasks completed and checked - ALL tests passing. Don't stop until you're done" and yet it keeps stopping anyway. 110
Steve Gordon @stevejgordon.bsky.social · 18/08/2026Last week we launched my latest course on @pluralsight.bsky.social - "Applying Dependency Injection & Middleware in ASP․NET Core 10" Please check it out at www.pluralsight.com/courses/appl... #dotnet #aspnetcorepluralsight.comApplying Dependency Injection & Middleware in ASP.NET Core 10 010
Steve Gordon @stevejgordon.bsky.social · 04/06/2026Helpful, thanks @visualstudio.com. What is "s"! Had to kill the process. I feel it may be punishing me for opening it less now I'm mosting in AI terminals and VS Code! #visualstudio 040
Steve Gordon @stevejgordon.bsky.social · 02/06/2026At the gate and soon on my way to NDC Copenhagen hosted by @ndcconferences.com. Hope to see a few familiar faces and meet many new ones while I'm there! 010
Reposted by Steve GordonNDC Conferences @ndcconferences.com · 28/05/2026NDC Copenhagen is almost here, and we couldn't be more excited! If you haven't grabbed your ticket yet, there's still time. Come join us next week for world-class talks, brilliant speakers, and great company. See you there! 🇩🇰 Tickets -> ndccopenhagen.com #NDCCPH 041
Steve Gordon @stevejgordon.bsky.social · 26/05/2026So GitHub Actions is down (again) and blocking me! Great! 000
Steve Gordon @stevejgordon.bsky.social · 20/05/2026I built in Azure as a) I wanted to build some experience there b) MVP contacts c) its a genuinely good fit and d) I'm using Aspire to the dev story is pretty nice. Just a shame that when I come to deploy, I can't do it where I want! 100
Steve Gordon @stevejgordon.bsky.social · 20/05/2026It depends on other services, key vault, service bus, blob storage too. I'm not against alternatives. I have a lot of prior AWS experience. However the app is written and working with the Azure libraries already so a switch is non trivial extra work. 100
Steve Gordon @stevejgordon.bsky.social · 20/05/2026I think its general computer constraints so I'm not sure. Also the app is written so it would be a pain to have work to switch. I'm mostly going to have to consider hosting elsewhere if other regions are not suitable. 000
Steve Gordon @stevejgordon.bsky.social · 20/05/2026The data this is planned to hold may have to reside in the UK which is problematic! I'm going to assess whether EEA is suitable, it's just a pain. Also doesn't build confidence that if i need new services later I don't hit constraints elsewhere. 000
Steve Gordon @stevejgordon.bsky.social · 20/05/2026Pretty staggered to find that resource constraints in Azure mean I can't deploy a side project requiring SQL to any UK region currently. The promise of the cloud model is somewhat failing at this point and I have to assume that AI is crippling the availability of compute. 420
Steve Gordon @stevejgordon.bsky.social · 15/04/2026This is the area where I'd really struggle myself to build out the UI any better. Again, fine for MVP/POC but you'd want an experienced designer and ux person to improve everything for a real product launch. 020
Steve Gordon @stevejgordon.bsky.social · 15/04/2026UI-wise, it started quite well. The original designs were generic and basic, but I spent time in several design sessions guiding it to think out of the box and try 5 different designs. Then I'd pick the bits I liked and keep refining until we had something reasonably decent. 110
Steve Gordon @stevejgordon.bsky.social · 15/04/2026From a output perspective, it's a mixed bag. The backend razor pages application code works but isn't how I'd write it. For building a MVP/POC it's kinda okay, but it would need refactoring before launch. It's EF models are not always that sensible so it needs some handholding. 110
Steve Gordon @stevejgordon.bsky.social · 15/04/2026Even with subagents it doesn't seem to leverage them particularly well if I leave it to it. 110
Steve Gordon @stevejgordon.bsky.social · 15/04/2026Workflow wise... not great. I'm back to manually launching a specific agent profile and guiding it through each stage, planning, implementing, testing, manually. The recent 5 hr session limits have made my planned flow impractical. It burns too quickly. 110
Steve Gordon @stevejgordon.bsky.social · 15/04/2026The cynic in me thinks this is a great strategy to be intentionally dumb and burn my allowance. It used to handle this just fine. On top of that it's ignoring my skill guiding it how to run TUnit tests, and I watch it try something like 8 times with an invalid command. So it's also wasting time. 000
Steve Gordon @stevejgordon.bsky.social · 15/04/2026Sonnet 4.6 is annoying me! Gave it a clear openspec with 8 deliverables. Asked it to implement deliverables 1 to 3 (more token efficient) but it implements randomly tasks from all deliverables, but not all tasks for all deliverables, burning a huge chunk of my #claudecode pro 5 hour window! 110
Steve Gordon @stevejgordon.bsky.social · 09/04/2026I find the sessions very useful to refine the idea and I often ask them to come up with 10 "outside of the box" ideas to consider. Often this teases out some useful thoughts we can further iterate on until we have a really interesting and detailed plan for the feature that we can then spec. 010
Steve Gordon @stevejgordon.bsky.social · 09/04/2026One of the things I enjoy about AI agents is the brainstorming phase. I have CEO, Requirements Analyst, PM and Solution Architect personas that I use to take a feature idea and have them challenge me on the functionality and technical aspects. I ask them to be 95% confident my intent is understood. 100
Steve Gordon @stevejgordon.bsky.social · 06/04/2026I've been using Sonnet 4.6 and so far I'd say GPT 5.4 is easily comparable in quality. I'm building from very refined specs though. It seems faster and seems to faff less to achieve the goal. I'm interested to see how it does on a more complex, open ended problem. Opus may still be the one to beat. 000
Steve Gordon @stevejgordon.bsky.social · 06/04/2026Codex has got me back to productive on my side project this morning. Churning through deliverables on my next feature around preparing the children to head out to a local zoo with the occasional prompt. Claude usage was 77% 5hr after the first small change. Codex is on its 10th with 65% remaining. 020
Steve Gordon @stevejgordon.bsky.social · 05/04/2026I'm on PTO for a week but when im back, I'm interested to see the burn rate on work stuff using the Enterprise plan. I hit $55 in one day and bar one refactoring, most tasks were smaller and focused on a small number of CI and build files + some tests. Seemed excessively high for what i achieved. 000
Steve Gordon @stevejgordon.bsky.social · 05/04/2026Same. Before the last week or so I'd been planning to upgrade to max for a side project but it's a no go until trust is won back that it'll be worth the personal investment. Codex is working out for now and £20 plan is enough to unblock progress while I wait onto see how this shakes out. 010
Steve Gordon @stevejgordon.bsky.social · 05/04/2026Ive been pure Sonnet and still hitting massive usage. I started a new session with a basic prompt and after one file read i was at 18% of the 5 hr usage. Gave up after that. 000
Steve Gordon @stevejgordon.bsky.social · 05/04/2026And the results are in. For the next deliverable, same basic refactor against three files (3 tasks), Codex with GPT5.4 medium used 4% of my 5hr window. 000
Steve Gordon @stevejgordon.bsky.social · 05/04/2026I used to be able to work through a 1, usually 2 complete openspecs for new features in the 5hr window (Pro). Now I'm struggling to get 3 deliverables (3 of 11) from one change. The last one burned 26% for 3 small tasks touching two files. 100
Steve Gordon @stevejgordon.bsky.social · 05/04/2026I'm trying out Codex CLI and the pro subscription this morning. Claude Code seems to have become unusable with the whole token usage fiasco. I've removed most of my helpful infrastructure to save tokens but even plain openspec apply tasks are burning my usage and the changes are slow too. 520
Steve Gordon @stevejgordon.bsky.social · 27/03/2026Another example of AI acting like a lazy developer (and yes, humans do this too). Here @github #copilot decided to ignore common file separation and use "a shortcut" by just throwing them into an existing test file. Overall it saved me time on a chore, but I still have to watch closely. 020
Reposted by Steve GordonILSpy - .NET Decompiler @ilspy.bsky.social · 27/03/2026The final preview of #ILSpy 10 is ready for download github.com/icsharpcode/... 0137
Steve Gordon @stevejgordon.bsky.social · 26/03/2026I'm continuing to experiment with subagents and an automated ochestrated workflow for feature implementation, with QA and code review. It's kinda working, but it burns tokens and several steps just fail and the ochestrator (main session) end up doing the work again. Needs more investigation! 010
Steve Gordon @stevejgordon.bsky.social · 25/03/2026I thought I had the dotnet-data plugin installed, but it seems not on this machine. Maybe (hopefully) that would have helped here. 000
Steve Gordon @stevejgordon.bsky.social · 25/03/2026One of the reasons I'm spending time learning about #ClaudeCode subagents is that while my workflow has been quite productive, I'm becoming the slowest part of the process, manually orchestrating various sessions. Keeping track became hard and tiresome so I ended up with notes to keep track! 110
Steve Gordon @stevejgordon.bsky.social · 25/03/2026This morning I am playing with #ClaudeCode subagents and building a test workflow. 030
Steve Gordon @stevejgordon.bsky.social · 25/03/2026Yuk! Caught Claude Code using AsyncLocal to work around a pooled EF DbContext rather the recommended pattern of scoped factory. Code reviews are still important people! The challenge is that I find reviewing/reading code I haven't written quite slow as I need to build up the context. #dotnet 172
Steve Gordon @stevejgordon.bsky.social · 20/03/2026One unhealthy habit introduced with my personal Claude Pro subscription is wanting to maximise my usage. I now find myself starting as session as soon as I wake to maximum the number of 5 hour windows I get. Also juggling prompts to get the maximum 5hr and weekly token usage. 230
Steve Gordon @stevejgordon.bsky.social · 14/03/2026I did have a hook set up on stop (which may have been the wrong place) you run my devlog skill which include the note about prompting for a commit decision. Ive removed the hook and the skill seems to work fine on its own now. Any good resources for proper hook usage? 000
Steve Gordon @stevejgordon.bsky.social · 14/03/2026I must admit, for my current learning project I have been. My goal was to see if I could get something built with limited input from me while on holiday. I'd start a prompt and let it fly so I was using dangerously skip permissions. For "real" dev during a work day I would not do that!! 000
Steve Gordon @stevejgordon.bsky.social · 13/03/2026Pembrokeshire. Some beautiful coastline to explore. Yeah, TBF its our fault for buying an XC90 with huge damn wheels (wasn't easy to change on an uneven pull in on a country lane). It's the first time I've had to buy a new tire so the price surprised me too!! 000
Steve Gordon @stevejgordon.bsky.social · 13/03/2026At a motorway services on our way home from 11 days in Wales. It's been a nice break with lots of activities and outings for the girls. Had a surprise extra cost of £313 yesterday after bursting a tire in a nasty hidden pothole. Glad we could get it replaced rather than limp home on the spare! 100
Steve Gordon @stevejgordon.bsky.social · 13/03/2026On one hand it's just somewhat frustrating but on the other it sets a worrying precedent about trust. If it can follow the claude instructions properly, is it following my prompts fully? I have also found its working around my git settings to avoid signing!! 000
Steve Gordon @stevejgordon.bsky.social · 13/03/2026My current project is not public but at some point I plan to share my workflow once I've refined it a bit further. Likely as a blog post and perhaps as a repo with the skill in it. 011
Steve Gordon @stevejgordon.bsky.social · 12/03/2026I've been refining this workflow and its now quite efficient. Agents log at the end of a session and include a handover (if needed) to another persona. I can then quickly bring a new session up with the correct role and point it at the handover to continue the next steps. 100
Steve Gordon @stevejgordon.bsky.social · 11/03/2026In a side project experiment with AI, mostly #claudecode, I've been reviewing the code its written before committing but I was kinda ignoring the tests. Now having reviewed some of them, they need some major work to remove brittle asserts and teach it about parameterised TUnit tests!! 130
Steve Gordon @stevejgordon.bsky.social · 11/03/2026It's kinda working but needs some refinement. Going to also look at hooks etc as a way to automate it more. Ideally though its like it to work across agents, cursor, github copilot too. Not finding a super DRY way to do that yet. Will share more if I bend it to my will! 020
Steve Gordon @stevejgordon.bsky.social · 11/03/2026After a bit of an infuriating time bouncing between #claudecode sessions in a dev and then qa persona, looping until it finally does things reasonably well, I've started experimenting with a devlog skill to summarise what is done, left over in each session and provide a handover the next agent. 110
Reposted by Steve GordonDylan Beattie @dylanbeatt.ie · 10/03/2026It's Claude Time again! Today @rendle.dev and I are joined by @richcampbell.bsky.social - let's see if Claude Code can build user interfaces? Can it understand sketches and scribbles? Tune in: twitch.tv/dylanbeattie (or youtube.com/dylanbeattie if Twitch isn't your jam)twitch.tvTwitchTwitch is the world 181