morisil

5 ms·

1.
by morisil 3mo ago
I am shocked how much my experience is different from yours. I wrote Claudine, my own version of Claude Code, almost 2 years ago. This experience gave me the understanding of how the technology works. Since then I've produced maybe 300
2.
by morisil 4mo ago
Quite aligned with my own experience from harness engineering and winning AI4Science hackathon. During the hackathon I was working as a human optimizer, moving the feedback from test harness running on Claude Code, back to my local Claude C
3.
Show HN: Markanywhere – A Streaming Processor of Meanings(github.com)
2 points by morisil 4mo ago|0 comments
4.
by morisil 4mo ago
I've been teaching harness engineering for a while: https://xemantic.com/ai/workshops/
5.
by morisil 4mo ago
How do you handle possible prompt injection in emails?
6.
by morisil 4mo ago
https://github.com/xemantic/markanywhere Incremental Markdown parser that emits streams of semantic events, plus tools to manipulate them - designed for real-time rendering of streamed LLM output.
7.
by morisil 5mo ago
Regarding applications/industries - after software development as an industry I see scientific research benefiting the most from agentic AI and autonomous machine reasoning. Another domain which is still not saturated would be somethin
8.
by morisil 8mo ago
More than a year ago I built my own coding agent called Claudine. I also made agentic anthropic-sdk-kotlin and few other AI libraries for the ecosystem. Maybe this low-level experience allows me to use these tools to deliver in 2 days what
9.
by morisil 9mo ago
Suboptimal choice. According to AutoCodeBench, for equivalent problem complexity, LLMs generate correct Kotlin code ~70% of the time versus ~40% for Python, and Go scores lower than Python. Kotlin can be executed as a script while providing
10.
I just wrote 'Global Agentic Night Berlin we are cooking'(buttondown.com)
1 points by morisil 10mo ago|0 comments
11.
Show HN: Neo4j SDK with minimal cognitive load for an LLM
1 points by morisil 10mo ago|0 comments
12.
by morisil 1y ago
1. LLMs excell at extracting facts from the context. Storing them as a subject-predicate-object relationships is "natural" for graph databases. Doing it right, so that this knowledge can be utilized more efficiently than any RAG,
13.
LiveCodeBench Pro – better than SWE-bench and aider's polyglot bench?(livecodebenchpro.com)
1 points by morisil 1y ago|0 comments
14.
by morisil 1y ago
I would support every political and social movement progressing us on the spectrum from patriarchy to matriarchy. In particular I would put pressure on the legal system of countries where women are still not allowed to inherit land and prop
15.
Ask HN: Qwen3 – is it ready for driving AI agents?
1 points by morisil 1y ago|0 comments
16.
by morisil 1y ago
Personally I use it for implementing complete new features using test driven development, or fix bugs in existing code base. Outside of software development I use it for managing all the legal documents of our cooperative (and it's Ger
17.
by morisil 1y ago
It is exactly the opposite. Participants collect all the pieces of knowledge to build an agent like my Claudine at the end of the workshop: https://github.com/xemantic/claudine/
18.
Ask HN: How to teach agentic AI? Please share your experience
9 points by morisil 2y ago|9 comments
19.
by morisil 2y ago
I added this functionality already some time ago in my Claudine agent: https://github.com/xemantic/claudine/ It costed roughly 30 lines of code: https://github.com/xemantic/claudine/blob&
20.
by morisil 2y ago
I am the author of Claudine, who presented the project during this meetup, and I am happy that it is inspiring others. Claudine is intended for educational purposes, so that anyone can easily build a powerful (e.g. Unix-omnipotent) autonomo
21.
Show HN: Cognitive Challenges of AI Code Generation Across Languages(github.com)
1 points by morisil 2y ago|0 comments
22.
I just wrote 'AI, workshops, the future '(buttondown.com)
1 points by morisil 2y ago|0 comments
23.
by morisil 2y ago
Please also take a look at my omnipotent Claudine. It is good at self-modifying to develop new tools: https://github.com/xemantic/claudine/
24.
Show HN: Parallax scrolling shader in infinite non-Euclidean refractive space(xemantic.com)
4 points by morisil 2y ago|0 comments
25.
by morisil 2y ago
Is there anything special required for brainstorming except for system prompt? You can try providing new system prompt to my Claudine: https://github.com/xemantic/claudine/
26.
by morisil 2y ago
I am a shader developer, often struggling with the incompatibility of underlying rendering tech (OpenGL, Vulkan, Metal, etc.). I built my own live-coding visual synthesizer using Kotlin script and OpenGL. If I had some time, I would port it
27.
by morisil 2y ago
I am happy with JReleaser, I just started migrating all my libraries from publishing through io.github.gradle-nexus.publish-plugin to publishing with jreleaser, which became necessary after migrating my sonatype account to the new maven cen
28.
Show HN: Power-assert for hierarchical data structures(github.com)
2 points by morisil 2y ago|0 comments
29.
by morisil 2y ago
Location: Berlin Remote: Yes Willing to relocate: No Technologies: Agentic AI, LLMs, Kotlin/Multiplatform, Java, GLSL/Shaders, Graphics/GPU programming, JavaScript, Spring Boot, DevOps, AWS, GCP, BigQuery, PostreSQ
30.
by morisil 2y ago
SEEKING WORK | Berlin | Remote Technologies: Agentic AI, LLMs, Kotlin/Multiplatform, Java, GLSL/Shaders, Graphics/GPU programming, JavaScript, Spring Boot, DevOps, AWS, GCP, BigQuery, PostreSQL, Oracle, ETL, Terraform, softwa

More ›