🔭 Contributing to OpenSRE — an open-source AI SRE benchmark
🛠 Software Engineer @ Caterpillar · based in Chennai, India
🤖 Curious about agentic systems & production incident response
🌱 Learning Cloud + DevOps with a Python-first lens
⚡ Clean code, fast feedback loops, evidence over hunches
🔭 OpenSREAn open-source AI SRE benchmark + agent runtime. Turns a raw alert into a structured root-cause analysis by running an LLM ReAct loop against 60+ observability, cloud, database, and incident-management tools. Investigations run on your own infrastructure — bring your own LLM (Anthropic, OpenAI, Bedrock, Ollama, …). My contributions: terminal-renderer test coverage (#2606, merged), Dockerfile build regression suite (#2620), synthetic test analysis (#2599). Stack: Python · FastAPI · Docker · Terraform · AWS · Postgres · Redis |
A harness-native operator system for agentic work with Claude Code. Ships ~20 language-specific rule packs (Python, TypeScript, Go, Rust, Java, Kotlin, Swift, C++, C#, Ruby, PHP, Dart, F#, Perl, ArkTS, Angular, web…), 50+ agents, slash commands, hooks, and platform configs. My customizations: tuned for Python / web / agentic SRE workflows, with hooks and rule overrides for my own day-to-day. Used daily across my open-source contributions (OpenSRE and others). Stack: Shell · TypeScript · Python · Markdown · npm — based on ECC by @affaan-m, MIT |
The snake needs a GitHub Action to regenerate daily. Add
.github/workflows/snake.ymlinmaheshsv/maheshsv(see the file in this repo). The image will 404 until the first workflow run completes — trigger it manually via the Actions → Generate Snake → Run workflow button after committing.