Explore Prompts

Create a calibrated confidence score with abstain rules when evidence is thin or conflicting. Provide messaging strategies and re-query tactics.

Write a polite crawling policy: robots.txt compliance, crawl budgets, backoff, and legal safe-use notes. Provide telemetry fields to prove compliance.

Define a plugin interface for new corpora with auth, rate limits, pagination, and schema normalization. Provide example adapters and a certification checklist.

Add a privacy stage that redacts PII from inputs/outputs and blocks storage for sensitive fields. Provide audit logs and consent flags in metadata.

Design a DAG (Airflow/Prefect) for reproducible research: crawl→retrieve→synthesize→verify→export. Include artifact hashing and cache invalidation rules.

Create an exporter that converts answers into JSON/CSV with fields (claim,evidence,source_url,confidence,timestamp). Provide schemas and sample exports.

Specify table extraction, PDF figure capture, and caption grounding. Define when to quote vs paraphrase, and include a screenshot-to-citation workflow.

Design watchers for tracked topics that diff sources and trigger re-summaries when material changes occur. Provide noise filters and notification policy.

Create a pipeline that generates questions from documents, with required answer spans and citations. Provide filters for ambiguity and leakage checks.

Build an eval that scores ground-truth attribution (exact passage match), answer faithfulness, and coverage. Provide dataset schema and a nightly regression plan.

Propose a router that routes by budget and SLA: small model for recall, larger for synthesis, tools only when signal warrants. Provide policy examples and guardrails.

Create a memory layer that stores hypotheses, decisions, and unresolved questions. Provide schemas, TTL rules, and a notebook export (Markdown/CSV).