Explore Prompts

Design a user experience: question intake, clarifying questions, showing sources, evidence table, and exporting. Include interaction patterns for analysts vs casual users.

Design content versioning: snapshots, change logs, and references to “as of” dates for citations. Include retention and storage cost considerations.

Design incremental crawling: store ETags/Last-Modified, compute diffs, and update indexes incrementally. Include pitfalls and how to handle missing headers.

Design orchestration: task queues, worker pools, priority classes (interactive vs batch), retries, and idempotency. Include a simple architecture for cloud deployment.

Design caching: query result caches, page caches, and embedding caches. Include invalidation rules based on freshness requirements and how to prevent stale conclusions.

Build a cost model: per-query search API costs, per-page fetch cost, storage, embeddings, and LLM tokens. Provide formulas and a budgeting plan with guardrails.

Create a unified rate limit manager: per-provider quotas, burst control, adaptive throttling, and retry policies. Include how to schedule jobs to maximize throughput within limits.

Design security controls: API keys storage, least privilege, environment separation, and audit logging. Include rotation policy and incident response playbook.

Design PII handling: detection, redaction, retention limits, and access controls. Include how to avoid collecting unnecessary personal data and how to respond to removal requests.

Draft a safety policy for the research bot: sensitive topics, privacy boundaries, defamation risk, and when to ask the user for confirmation. Include a user-facing explanation style.

Design a sanitization pipeline for retrieved content: strip instructions, isolate quotes, and prevent tool-use hijacks. Include adversarial test cases and regression suite.

Create a QA workflow: sample-based review, check citations, verify key claims, and approve final briefs. Include reviewer guidelines and a severity rubric.