Rishikanth Chandrasekaran
Community@rishikanthc · Bellevue
Applied Researcher at eBay
Agent Skills by Rishikanth Chandrasekaran
Showing 20 vetted skills indexed across 1 GitHub repositories.
using-git-worktrees
Create isolated git worktrees with dependency setup and baseline test verification.
test-driven-development
Enforce failing tests before implementing production code across testing workflows.
ml-paper-to-code
Translate research paper equations into tested model implementations with shape checks.
systematic-debugging
Identify root causes of bugs and failing tests before applying fixes.
using-superpowers
Enforce Skill tool invocation and checklist-driven workflows before responding.
dispatching-parallel-agents
Dispatch parallel AI agents to investigate multiple unrelated failures independently.
ml-test-driven-development
Detect wiring and integration bugs in PyTorch model implementations before training runs.
executing-plans
Execute implementation plans in batches with dependency checks and checkpoint reports.
finishing-a-development-branch
Verify tests and guide git branch merge, push, preserve, or discard decisions.
ml-evaluation-framework
Enforce statistical evaluation and reporting for machine learning experiments.
brainstorming
Turn ideas into validated design documents through iterative questioning.
ml-code-standards
Enforce reproducibility and coding standards for machine learning codebases.
writing-plans
Create implementation plans with TDD tasks, file paths, and .tasks.json.
requesting-code-review
Review git commit ranges for code quality, architecture, and tests.
receiving-code-review
Evaluate code review feedback and produce a prioritized implementation plan.
ml-training-pipeline
Set up distributed PyTorch training pipelines with DDP, bf16, and checkpointing.
writing-skills
Create and verify SKILL.md documentation using RED-GREEN-REFACTOR pressure scenarios.
ml-huggingface-models
Verify, download, fine-tune, and integrate HuggingFace pre-trained models in PyTorch projects.
verification-before-completion
Run validation commands and inspect outputs before claiming completion.
subagent-driven-development
Dispatch fresh subagents per task with two-stage reviews in a single session.