Operator-grade prompts for Intent Solutions engagements. Every asset mirrors our service lines—Private AI, AI Agents, Automation, Cloud & Data, and Learn—and is written for production operators, resellers, and execs who expect receipts.
🌐 Browse the live catalog
📬 Start a project
| Service Line | Flagship Prompt | What it Delivers |
|---|---|---|
| Cloud & Data | 15-ops-devops-system-analysis |
The definitive operations guide for a platform hand-off. |
| Cloud & Data | 15-ops-github-release-runbook |
Release documentation, deployment plan, and communications pack. |
| Cloud & Data | 03-cloud-security-hardening |
Prioritised hardening backlog mapped to compliance frameworks. |
| Automation | 01-auto-n8n-flow-upgrade |
Resilient n8n workflow blueprint ready for reseller packaging. |
| Automation | 01-auto-incident-escalation |
Automation-assisted escalation matrix with communication scripts. |
| AI Agents | 01-agent-copilot-design |
Claude copilot design brief tied to operator workflows. |
| AI Agents | 01-agent-slash-command-draft |
Slash-command bundle aligned with this library’s standards. |
| Private AI | 01-ai-private-model-rollout |
Safe, observable rollout plan for private models on Vertex/Bedrock. |
| Learn | 12-learn-operator-onboarding |
Two-week enablement curriculum for new operator teams. |
| Learn | 12-learn-master-doc-filing |
Chronological documentation discipline for every deliverable. |
| Learn | 12-learn-prompt-retrospective |
Workshop to measure, tune, and sunset prompts responsibly. |
See the full index in prompts/README.md or explore via the GitHub Pages site above.
---
id: unique-slug
title: Human-readable name
service_line: cloud-data | automation | ai-agents | private-ai | learn
audience: primary reader
intent: one-sentence job to be done
last_reviewed: YYYY-MM-DD
model_hint: claude-3-5-sonnet
tone: descriptor
delivery: expected deliverable format
---- Pick the prompt that matches your engagement or slash command.
- Fill in variables like project name, stakeholders, or severity levels.
- Feed into Claude (or the approved model) and iterate with evidence.
- Capture outputs in
01-Docs/using the MASTER DIRECTORY STANDARDS. - Keep numeric prefixes (
01-,03-,12-,15-) intact so sorting stays consistent across tooling. - Archive variants to
99-Archive/legacy-prompts/once superseded and note the change in an AAR.
- Duplicate the schema above and align to the proper service line.
- Keep instructions operator-first; describe decisions, not shell commands.
- Reference real artefacts (file paths, dashboards, tickets) so outputs are verifiable.
- Update
prompts/README.md, runpython tools/generate_prompt_catalog.py, and add the card to the site if needed. - Include before/after screenshots or summaries when opening a pull request.
This repository follows MASTER DIRECTORY STANDARDS:
- Documentation lives under
01-Docs/withNNN-CC-ABCD-description.extnaming. - Prompt library is flat under
prompts/with kebab-case file names. - Superseded assets move to
99-Archive/legacy-prompts/with an associated after-action report.
MIT License — see LICENSE.
Intent Solutions · Operator-first AI systems · intentsolutions.io · startaitools.com