Prompt engineering

Notes on prompting as an engineering interface: capability boundaries, evidence contracts, and operational templates.

This hub maps Explanation (articles) → How-to (procedures) → Prompt templates → Policies (normative rules), aligned to an intent-based docs split.

Start here

Prompt Engineering Guide for Daily Work (Deep Dive)
A “daily work” prompt workflow (concept + examples)
Next: Prompt Engineering Guide for Daily Work
Objective Technical Baseline Rules (No Simulation) — policy
Baseline non-simulative constraints (instruction + boundary discipline)
Next: /prompts/objective-technical-style-non-simulative.system.txt · /prompts/instruction-hierarchy-and-evidence-boundary.system.txt
Choose allowed sources for factual answers
Choose an evidence boundary (artifacts-only vs authoritative)
Next: Facts-only: Artifacts-only (no external sources) · Facts-only: Authoritative sources required (citations required)
Facts-only: Authoritative sources required (citations required)
A hard refusal/evidence contract for factual claims (with citations)
Next: /prompts/facts-only-authoritative-sources-required.system.txt · Facts-only: Authoritative sources required (citations required) — procedure
Web Verification & Citations Policy
Web verification + citations (tooling + output discipline)
Next: /prompts/web-verification-and-citations.system.txt · /prompts/web-browsing.user.txt · Request web browsing (prompt template)
Chain-of-Verification — policy
A structured verification loop (pre-output self-check)
Next: /prompts/chain-of-verification.system.txt · Run the Chain-of-Verification (CoVe) — procedure

Pages in this section

Core pages for this topic are listed above under “Start here”. Additional pages will appear here as they’re published.

Section resources

Related Related