Analyze the shared canonical literature and repo library for a specified field or direction, summarize current trends, propose candidate research programs before a concrete program exists, and list relevant papers or repos with basic metadata and short summaries. Use when Codex needs to choose a direction, control scope from macro to micro, or browse by tag/topic without ingesting new sources or mutating canonical entries.
Use this before research-conductor when direction selection is still open and llm-wiki landscape synthesis is needed.
ingest (not owner): does not import new raw sources.query (owner): run field-scoped retrieval across canonical literature/repos and tag taxonomy.lint (limited): validate landscape report schema and seed program completeness.index (limited): maintain discoverable landscape artifacts under library/landscapes/.log (owner): persist survey assumptions, scope mode, and ranking rationale.research-conductor) and program-scoped deep evidence maps (literature-analyst).tags, , titles, and .topicsshort_summarykb/library/landscapes/ with trend summaries, candidate program seeds, and conductor-ready prompts.research-conductor via create-program-from-landscape.research-deliverable-curator instead of treating the landscape summary as a navigation page.macro, focused, or micro modes.kb/memory/domain-profile.yaml.survey_id plus program_seed_id.library/landscapes/, not left as chat-only recommendations.python3 .agents/skills/research-landscape-analyst/scripts/survey_landscape.py survey --field "your target field" --scope focused
python3 .agents/skills/research-landscape-analyst/scripts/survey_landscape.py survey --field "your target field" --scope macro --tag tag-a --tag tag-b
python3 .agents/skills/research-landscape-analyst/scripts/survey_landscape.py list --kind literature --tag tag-a --limit 10
python3 .agents/skills/research-landscape-analyst/scripts/survey_landscape.py list --kind repos --query "your field keyword" --limit 5
python3 .agents/skills/research-conductor/scripts/manage_workspace.py create-program-from-landscape --survey-id my-field-scan --program-seed-id seed-1
kb/programs/<program-id>/ here.literature-analyst for program-scoped synthesis once a concrete program exists.research-deliverable-curator.skill-evolution-advisor.