Run blueprint-driven research workflows with FAROS
Bind a research profile to a FAROS blueprint and execute an AutoResearch chain from idea generation through experiment scaffolding, paper drafting, and reviewer simulation.
npx skills add agentskillexchange/skills --skill run-blueprint-driven-research-workflows-with-faros
pip install -r requirements.txt, install frontend dependencies with npm install, start the documented services, then run a FAROS request with the ml_paper blueprint and the desired research profile.Use FAROS when the job is to run a structured research-production workflow with artifacts at every stage, not to chat with a single AI scientist prompt. The operator starts the backend, binds a profile to the `ml_paper` blueprint, runs the current `idea -> experiment -> paper -> review` chain, and collects code project scaffolds, experiment records, venue-aware LaTeX paper artifacts, PDFs, and structured review reports. The useful agent behavior is to treat the blueprint as the contract, preserve the generated artifacts, and inspect the review output before deciding whether the research packet is ready for another iteration. Invoke this instead of using a research assistant normally when the workflow needs blueprint constraints, capability adapters, profile/provider binding, event logs, artifact persistence, and a repeatable review stage. The boundary is the FAROS AutoResearch runtime and its documented LLM-research blueprint; it should not be stretched into a generic paper-writing app or all-purpose research framework. Keep iterations tied to the documented blueprint outputs.