Prysai Lab · Repositories · Core projects
I build and inspect AI systems that hold up beyond the demo: agents, retrieval, evaluation, and tool use in real developer workflows.
I follow the full agent engineering loop:
- choose the right model and context
- connect tools safely and make behavior observable
- evaluate quality, regressions, and failure modes
- turn useful experiments into repeatable workflows
The goal is simple: fewer impressive demos, more systems that can be tested, understood, and improved.
Prysai is the home for product experiments and practical work around AI systems and developer tools. The emphasis is on small, testable pieces that can be inspected, measured, and reused.
Six upstream projects, seven merged pull requests — documentation, packaging, compatibility, performance, privacy, and UX fixes that are now part of other projects.
| Project | Contribution | Merged PRs |
|---|---|---|
| MSKazemi/kubeintellect | Documented v4 data-handling boundaries and updated the kube-q Homebrew formula with current release resources. |
#105 · #111 |
| mastra-ai/mastra | Fixed CommonJS schema compatibility by bundling the Zod conversion dependencies and adding regression coverage. | #21147 |
| squid-protocol/gitgalaxy | Replaced manual append loops with extend-based recorder code and resolved the related Ruff performance findings. |
#1201 |
| abenneto/gandharva | Refined README formatting and section headers. | #19 |
| seva9523/EvalRepro | Added --no-id-preview to suppress sample ID previews while preserving sample and field digests. |
#12 |
| felladrin/MiniSearch | Capped staggered result animations and honored prefers-reduced-motion. |
#2324 |
All entries above are public PRs confirmed as merged into their upstream repositories.
A compact working set of upstream projects I currently study, use, or contribute to. They are selected for technical leverage, not repository size or star count.
|
RAGOps Evaluation infrastructure Offline regression tests and explainable release gates for RAG systems and AI agents. upstream · working fork |
PydanticAI Typed agent runtime Explicit dependencies, tools, structured outputs, and validation for production-minded agents. upstream · working fork |
|
FastMCP Tool connectivity A practical protocol layer for building and connecting MCP servers and clients. upstream · working fork |
Promptfoo Quality and safety gates Evaluation, comparison, and red-teaming for prompts, agents, and RAG pipelines. upstream · working fork |
|
Browser Use Real-world execution Browser control as a practical interface for tool-using agents and workflow automation. upstream · working fork |
Spec Kitty Developer workflow Spec-driven development, worktrees, and repeatable workflows for serious software projects. upstream · working fork |
These links point to upstream projects and my working forks. They are study and contribution paths, not claims of original ownership.
This tracks the total stars received by my public repositories over time. It is an account-level view, not the star count of any single project.
All public repositories · forks count only their own stars · updated daily by GitHub Actions.
AI Agents · RAG · LLM Evaluation · MCP · Developer Tools · Codex Workflows
Building in public, one reliable workflow at a time.