architect
01soundwaverohit/quantum-research-hub
Agent Claude Code
Designs implementation plans and proposes repo improvements as review items. Never merges or modifies core safety logic directly.
Local-first autonomous quantum computing research lab powered by MCP, Claude, and Streamlit
soundwaverohit/quantum-research-hub
Agent Claude Code
Designs implementation plans and proposes repo improvements as review items. Never merges or modifies core safety logic directly.
soundwaverohit/quantum-research-hub
Agent Claude Code
Scores and ranks ingested papers by relevance, novelty, feasibility, and risk; assigns a recommended action. Strict by default.
soundwaverohit/quantum-research-hub
Agent Claude Code
Builds and improves the Streamlit Quantum Research Hub dashboard. Reads SQLite + artifacts; works even when the MCP server is not running.
soundwaverohit/quantum-research-hub
Agent Claude Code
Builds reproducible experiment folders from ideas — baseline, config, tests, metric logging, seed control, report scaffold.
soundwaverohit/quantum-research-hub
Agent Claude Code
Runs bounded local smoke experiments with strict timeout and approval gates. Saves logs and metrics. Needs approval for installs/GPU/long jobs.
soundwaverohit/quantum-research-hub
Agent Claude Code
Converts ranked paper clusters into small, testable, source-cited research ideas. Prefers one small experiment over a grand claim.
soundwaverohit/quantum-research-hub
Agent Claude Code
Implements and tests the Researcher MCP server and its tool contracts — typed, documented, safe, budget-aware, structured errors.
soundwaverohit/quantum-research-hub
Agent Claude Code
Finds candidate quantum-computing papers from arXiv under the configured budget. Does not deep-read or generate ideas.
soundwaverohit/quantum-research-hub
Agent Claude Code
Builds compact paper cards from metadata + abstract (and retrieved chunks). Never feeds full PDFs to the model; never fabricates claims.
soundwaverohit/quantum-research-hub
Agent Claude Code
Writes concise daily and weekly research reports from papers, ideas, experiments, validator verdicts, and budget usage.
soundwaverohit/quantum-research-hub
Agent Claude Code
Writes and runs fast, deterministic pytest tests for core modules. Mocks the network; never relies on live arXiv or paid APIs.
soundwaverohit/quantum-research-hub
Agent Claude Code
Reviews experiments skeptically and returns accepted | rejected | inconclusive with a checklist. May reject results.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: