Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add MartinOlivero/saas-builder --skill architecture-primergit clone --depth 1 https://github.com/MartinOlivero/saas-builderWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/martinolivero/saas-builder/architecture-primer)<a href="https://agentmods.dev/skills/martinolivero/saas-builder/architecture-primer"><img src="https://agentmods.dev/badge/skills/martinolivero/saas-builder/architecture-primer/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/martinolivero/saas-builder/architecture-primer"><img src="https://agentmods.dev/badge/skills/martinolivero/saas-builder/architecture-primer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00111 | $0.00970 |
| Opus 5 | $0.00056 | $0.00485 |
| Sonnet 5 | $0.00022 | $0.00194 |
| Haiku 4.5 | $0.00011 | $0.00097 |
Grade A, and why
architecture-primer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 85 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Architecture Primer
This skill turns a vague "how should I build this?" into concrete, justified architecture decisions. It is based on the System Design Primer, adapted to favor shipping over over-engineering.
Analogy: do not design a 10-lane highway before knowing if 10 cars or 10 million cars will use it. Measure the traffic first, then build the road.
PHASE 1 — Discovery (mandatory before any decision)
Never propose an architecture before completing discovery. Ask questions 2-3 at a time, in plain language, in these three blocks. Skip any question the user already answered.
Block A — Scale
- How many users do you expect at launch? And in 12 months?
- Are users in one region or several countries?
- Is usage steady, or does it spike (e.g. launches, business hours, campaigns)?
Block B — Nature of the system
- Does it read more than it writes, or the other way around?
- What happens if it goes down for 1 hour — is it critical?
- Does the data need immediate consistency, or is a short delay acceptable?
Block C — Context
- Is there a technical team, or just you?
- Roughly what monthly infrastructure budget is realistic?
- When do you need the MVP live?
PHASE 2 — Decision tree
Apply these rules using the discovery answers. Each rule has a default; deviate only with a concrete reason.
Monolith vs Microservices
- Team under 5 people → monolith, always.
- MVP or under 6 months to ship → monolith, always.
- Over 100k daily active users → evaluate microservices for the parts that actually hurt.
- Rule: start monolith, split when it hurts. Premature microservices add ops cost with no benefit.
SQL vs NoSQL
- Relational data needing integrity → PostgreSQL.
- Flexible documents, horizontal scale → MongoDB.
- Cache / sessions / real-time → Redis.
- Full-text search → Elasticsearch.
- Rule: PostgreSQL by default; switch only with a concrete reason.
Cache
- Same query repeats a lot → Redis.
- Data that rarely changes but is read often → CDN or Redis.
- Data unique per user on every request → no cache (it would not hit).
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 85 lines · 111 tokens per session scan A 00e3ca5fce80
architecture-primer is a skill published in the GitHub repository MartinOlivero/saas-builder (2 stars, last pushed 22d ago), licensed MIT. It adds 111 tokens to every session and 970 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
architecture-paradigm-microkernel
Applies microkernel architecture with minimal core and plugin extensibility. Use when building platforms where third parties extend core functionality.
plan
Use when a task needs structured planning before implementation. Supports --delegate and round=N[,M].
preplan
Use when a problem needs clarification and agreement before planning begins. Supports --deep and --delegate.
ralph
Use when implementing a plan or executing a prompt that requires verified completion.
pathfind
Use when you have problems but don't know what to build.
analyze
Use when deep investigation is needed — project structure, requirement gaps, or root cause diagnosis. Supports --delegate.