Hi, I'm Prajwal
AI engineer building agent tooling and applied ML, with three years of backend and full-stack at Opslyft.
hdprajwal.dev · LinkedIn · X
- Re-Evaluating Android Malware Detection: Tabular Features, Vision Models, and Ensembles (MDPI Electronics, 2026)
- pageshelf: self-hosted place to publish the HTML and Markdown pages your coding agent writes, all sharing one theme, docs at hdprajwal.github.io/pageshelf.
- gh-widgets: self-hosted SVG widgets for GitHub READMEs, live at gh-widgets.hdprajwal.dev.
- open-frame: make slides, carousels, and thumbnails by describing them to your coding agent, docs at open-frame.dev.
- emberd: sandboxing runtime that runs AI-agent tool calls inside Firecracker microVMs, docs at emberd.hdprajwal.dev.
- BaitBench: benchmark for how often LLMs fall for prompt injections hidden in job posts, live at baitbench.hdprajwal.dev.
- QuackCode: workspace for running several CLI coding agents side by side per project.
- codexpass: CLI that lets your other tools use the OpenAI credential your Codex login already saved, docs at codexpass.hdprajwal.dev.
- Agent-chaos: a tool that breaks LLM agents on purpose to check they recover correctly. Not public yet.
- Gitwise: CLI that turns staged diffs into commit messages and plain English into branch names.
- Multimodal Malware Detection: Master's thesis comparing static-feature and visual-bytecode malware classifiers.
- Guardscribe: real-time speech moderation that masks toxic spans in under 400ms.
things I built because I wanted to, not because anyone asked
- Lucida: trace any image onto paper through your phone's camera.
- Annoying UI: a gallery of intentionally hostile UI components.
Reach out on Twitter (@_hdprajwal) or email hdprajwal01 [at] gmail [dot] com.