Sean Ryan·May 11Your 2nd Brain via meowary OSS: use AI to consume One Note, Confluence, ADRs, Source code …note: meowary is not my own work — it is an OSS (Open Source Software) project by Andrew Vasilyev that I find really useful — this article…
Sean Ryan·May 5OpenCode Security Tip: Lock Down and Double-check the LLM ProviderOpenCode (https://opencode.ai) is a simple and powerful CLI tool to make the most of LLMs to plan updates to text (code but also markdown…
Sean Ryan·Apr 20AIAD Tooling: Adopt GSD Lite for the best of GSD (Get Shit Done) with Minimal ProcessGSD is a popular GitHub repo designed to support AI-assisted development (AIAD) through Spec-Driven Development (SDD). Its key…
Sean Ryan·Apr 20AIAD Tooling: Comparing OpenCode vs Claude SpecKit vs GSDIf you’ve spent any time building apps with LLMs, you’ve probably hit the same wall: the model is capable, but your workflow is not.
Sean Ryan·May 16, 2025Automatically Evaluate Hostile Prompt Handling for your LLM Application [LLM Security Series — Part…Besides taking defensive measures and following best practices to secure your LLM application (see my Medium article Security Risks and…
InDevSecOps & AIbySean Ryan·May 9, 2025Security Risks and Mitigations for your LLM Application [LLM Security Series — Part 2]Whilst adding an LLM to your application opens the door to the power of genai, it also opens the door to hostile users who can manipulate…
InDevSecOps & AIbySean Ryan·May 4, 2025LLM Hackers Can See Your Prompt [LLM Security Series — Part 1]When constructing an LLM application, for security reasons it can help to think of your system prompt as already visible to any user. This…
Sean Ryan·Mar 24, 2025Typed vs untyped languages: considering the case for a culturally-based preferenceSoftware developers often have a strong preference for either typed or untyped languages. Given that it can be difficult to objectively…
Sean Ryan·Mar 17, 2025Where is the hard evidence that ‘better’ languages produce less bugs?A seemingly reasonable opinion to hold in software developer circles, is that strongly typed languages result in higher quality code. More…
Sean Ryan·Mar 8, 2025Prompt Engineering: Using an intermediate structured format to improve the generative quality of a…The vast knowledge of an LLM can be a double-edged sword: if for example you need to generate a custom version of a public standard such as…