Copy any of these prompts into Claude or ChatGPT to start a guided conversation. The AI will interview you to gather the context it needs, then produce the output.
Documentation
Document a system architecture
I need to document our system architecture for the team.
Interview me about the main components, how they
connect, the infrastructure, authentication model, and
data storage. Then write architecture documentation
that a new team member could understand in their first
week — covering the system overview, component
descriptions, data flows, dependencies, and known
limitations.
Create a runbook
I need to write a runbook for an operational procedure.
Ask me what the operation is, what tools are involved,
permissions needed, what commonly goes wrong, and how
to roll back. Then write a step-by-step runbook that
someone could follow at 3am during an incident — with
verification steps and an escalation path.
Vendor evaluation
We're evaluating an AI platform for our organisation.
Interview me about our requirements — number of users,
use cases, data sensitivity, integration needs,
compliance requirements, and budget. Then create an
evaluation framework covering security, authentication,
compliance certifications, integration capabilities,
admin controls, pricing, and vendor stability. Include
specific questions to ask the vendor and red flags to
watch for.
We need to choose between AI platforms for our
organisation. Ask me about our size, industry, primary
use cases, existing tools (Google Workspace, Microsoft
365, etc.), security requirements, and budget. Then
compare the leading options (ChatGPT, Claude, Gemini)
across security, admin controls, integration,
strengths by use case, pricing, and limitations.
Recommend which fits and why.
Security and governance
AI security assessment
I need to assess the security implications of deploying
an AI tool. Ask me about the tool, how it's deployed,
what data it'll process, who has access, our current
security posture, and applicable regulations. Then
produce a security assessment covering data flows,
access control risks, data leakage vectors, compliance
gaps, and recommended mitigations.
Build an AI governance framework
We're rolling out AI tools and need a governance
framework. Interview me about our organisation, which
tools we're deploying, key concerns, existing policies,
and who oversees governance. Then draft a practical
framework covering approved tools, data handling rules,
approval processes for new use cases, monitoring,
incident response, training, and roles. Make it
something people will actually follow, not file away.
For platform comparisons, both Claude and ChatGPT have some inherent bias. Cross-reference vendor documentation for the most balanced view.
