Prepare a dated, source-backed research brief that preserves conflicting findings and uncertainty.
69%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
Fix and improve this skill with Tessl
tessl review fix ./test/fixtures/collective-learning/research-brief/SKILL.mdRead the task and each supplied source before drafting. Use only authorized sources. Use templates/brief.md for the deliverable. Run scripts/build_brief.py with the input JSON, output JSON and output Markdown paths when Python 3 is already available. Do not install tools or fetch additional material without ordinary authorization. When Node.js is already available, scripts/build_brief.cjs provides an alternative with the same three file arguments. Check timezone applicability before choosing it.
Normalize publication timestamps to ISO calendar dates with scripts/normalize_sources.py. Keep publication date, observation date and retrieval date distinct when they differ. Preserve source identifiers and links; cite every quantitative finding. Exclude sources explicitly marked superseded. If two current sources disagree, retain both values and their citations; explain their populations, time windows or measurement differences. Do not average incompatible measurements or choose a convenient winner.
Use scripts/citations.py to assemble the factual findings. It is a drafting helper, not an independent correctness verdict. Check the generated brief against the sources. Describe missing information, limitations and unresolved contradictions. Separate factual findings from recommendations; make recommendations conditional when evidence is weak. Include a concrete follow-up action, its owner and the question it should resolve. Return both brief.json and brief.md. Keep a generated brief in draft until reviewed.
If a source lacks a valid date or has an ambiguous identifier, ask for clarification or record that uncertainty rather than inventing a date or citation. Consult references/source-rules.md for these cases.
Bitterbot-AI/bitterbot-desktop
2d4d05d
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.