rohithkandula19
UserMasterless, terminal-native coding agent (Claude Code-style: reads, edits, runs code) for Claude, Gemini, Cerebras, Groq, and Ollama. Provider-agnostic with evals, memory, security hardening, 200+ plugins, and MCP. MIT.
Categories
Indexed Skills (9)
eng-review
Engineering review of a plan or change on six dimensions 0-10 with what a 10 looks like, then repair.
investigate
Root-cause a failure from a stack trace — trace to source, form hypotheses, then hand a repro to the fixer subagent.
review
Security and correctness review over a diff, delegating a structured pass to the reviewer subagent.
autoplan
Survey the code, draft a written plan, and get confirmation before editing anything.
design-review
Score a design or plan on five dimensions 0-10, say what a 10 looks like, then fix the weak ones.
office-hours
Force the clarifying questions out into the open and wait for answers before writing any code.
qa
Drive a QA checklist against a running app or URL, probing endpoints and reporting what actually happened.
retro
Turn a session's lessons into durable notes by appending them to RONIN.md, the project memory file.
ship
Release behind verify gates — run the tests, read the diff, then a staged rollout with a rollback plan.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.