hosting-decision
Skill0 starsChanged 50 days agoMITView on GitHub
Decide where a business application should be hosted (French/EU sovereign cloud, hyperscaler, vendor SaaS, or on-premise on the client's own servers) before a software project is priced. Use when the user asks "cloud or on-premise?", "do we need HDS / SecNumCloud?", "is our data exposed to the CLOUD Act?", "can we host on our own servers?", or when a specification or quote leaves hosting undecided. Produces a decision, its cost consequences, and the on-premise option priced separately.
In this repository6 other files it gives its agents.
- compare-offersSkill
- estimateSkill
- legacy-takeoverSkill
- review-quoteSkill
- review-specSkill
- vendor-questionsSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 9
If the user passed arguments, they are: $ARGUMENTS
- Distinctive: line 42
Check the certified scope of a provider, service by service: a provider can be certified for health data hosting (HDS) while some of its …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Hosting decision — settle it before pricing
If the user passed arguments, they are: $ARGUMENTS
Hosting changes the architecture, the recurring cost, the security clauses and the contract. It must be decided before vendors price the project, not discovered during it.
Language: answer in the user's language. Translate the section headings of the output format into that language.
Workflow
- Collect the facts (ask only what is missing): data categories (personal, HR, health, financial, industrial secrets), users and locations, availability needs, offline needs, existing infrastructure and IT staff, regulatory constraints of the sector, customer contractual requirements (some clients require EU-only or on-premise), budget for recurring costs.
- Walk the decision tree in
references/hosting-options.md. - Always present on-premise as a separately priced option when the user or their customers raise it — never dismiss it and never slip it in for free. Price it with its full cost: hardware, licences, backups (including an off-site copy), monitoring, security patching, the people who operate it, and renewal after 4–5 years.
- Output:
## Recommendation
Hosting model + one-paragraph rationale.
## Constraints that drove it
Data categories, regulation, customer requirements.
## Options compared
| Option | Fits constraints? | Setup cost drivers | Recurring cost drivers | Who operates | Exit / reversibility |
## On-premise (priced separately)
What it requires, cost items, and what the buyer must staff.
## Questions to put in the specification
Numbered.Principles
- Accounts in the buyer's name (cloud account, domain, backups), whatever the model.
- Check the certified scope of a provider, service by service: a provider can be certified for health data hosting (HDS) while some of its managed services (managed databases, AI APIs) are outside the certified scope.
- A provider headquartered in the US remains subject to US extraterritorial laws (CLOUD Act) even when data sits in an EU region. If that matters to the buyer or its customers, prefer an EU-headquartered provider, or a qualified offer (SecNumCloud) when required.
- AI features: say where inference runs and whether data is pseudonymised before it leaves the application.
- Backups are only real once a restore has been tested.
Maintainer
Maintained by JAIKIN — AI, Odoo and custom software, Strasbourg (France). This section is directory metadata, not part of the answer.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

