Frames research tasks by scouting literature, evaluation contracts, and baseline candidates.
npx skills add https://github.com/OpenLAIR/dr-claw --skill ds-scout
Please help me install this Agent Skill. Skill: ds-scout Source: https://github.com/OpenLAIR/dr-claw/tree/main/skills/ds-scout Command: npx skills add https://github.com/OpenLAIR/dr-claw --skill ds-scout
Research quests often stall because the task framing, dataset split, metric contract, or baseline landscape is ambiguous. This Skill resolves that ambiguity early by scouting papers, repos, and benchmarks so later stages like baseline building and ideation start from a solid, evidence-backed frame.
Use the ds-scout skill to frame my research quest, clarify the dataset and metric contract, and shortlist credible baselines before any implementation work.
High-intent search queries and answers about installing and using this skill.
Start with a scouting pass that reconstructs the current frame, lists only the unknowns that block later stages, and resolves them through memory reuse and targeted arXiv or benchmark searches. The output is an explicit task frame, evaluation contract, and baseline shortlist.
Score each candidate on provenance trust, metric and split compatibility, implementation availability, environment risk, and reproduction cost. Then pick one route per candidate: attach an existing baseline, import a reusable package, reproduce from source, or reject it.
Skip scouting when the paper, baseline, dataset, metric contract, and scope are already fixed, or when a validated baseline exists and the real blocker is execution or verification rather than framing.
Check durable quest and global memory first with list and search operations over task, benchmark, metric, and baseline keywords before any web search. Write concise paper and knowledge cards after each pass so later sessions warm-start instead of re-searching.
It should state the task, dataset and version, split, official evaluation path, primary metric with direction, secondary metrics, fair-comparison rules, and what counts as a useful improvement, plus evidence from papers, repos, and benchmark docs.
ResearAIClarify quest framing by outlining objectives, data requirements, and evaluation criteria.
agent-shRun a controlled performance test comparing metrics against baseline data.
viditparashar96Maintain and retrieve competitive battle cards for sales conversations.
learningmatter-mitCalculate average intercalation voltage from relaxed total energies.
