Skip to content

docs: articulate AI-infra fit and how Substrate compares#315

Merged
scttfrdmn merged 1 commit into
mainfrom
docs/ai-fit-and-comparison
Jun 4, 2026
Merged

docs: articulate AI-infra fit and how Substrate compares#315
scttfrdmn merged 1 commit into
mainfrom
docs/ai-fit-and-comparison

Conversation

@scttfrdmn
Copy link
Copy Markdown
Owner

Makes two arguments the README previously only asserted — explicit, grounded, and honest. Added to both the README and the docs Scope & Philosophy page.

1. Why it suits AI-generated infrastructure

AI generates IaC at volume inside generate → test → fix loops, which need:

  • Determinism — a flaky failure is a false signal an agent acts on, wasting fix cycles.
  • Free / fast / offline — runnable on every generated candidate, no account/spend/latency.
  • Cost as a machine-gradeable guardrailIntent{MaxCost} fails a deploy that would blow a budget (an AI has no bill-shock instinct).
  • Seedable failure paths — actually exercise the retry/poll/wait logic generated code claims to have.

2. How it compares

An honest trade-off table vs LocalStack / moto / Real AWS that includes where Substrate is weaker — it doesn't run your workload code, by design. Framed as "a different tier," not "better than," consistent with the scope boundary. The one deliberate ❌ links to Scope & Philosophy.

Replaces the old throwaway "LocalStack hides bugs" line. Renamed the differentiators heading Why SubstrateCore capabilities so the new subsection owns the "why."

Docs build clean (strict link check passes). Deploys to Pages on merge.

The README asserted 'test harness for AI-generated infrastructure' but never
earned it, and the only competitor mention was a throwaway LocalStack jab. Make
both arguments explicit, grounded, and honest — they share one root cause (the
API-surface scope boundary is both the differentiator and the AI-fit reason).

README + docs/scope.md:
- 'Why Substrate for AI-generated infrastructure': determinism (flakes mislead
  the loop), free/fast/offline (run on every generation), cost as a
  machine-gradeable guardrail (Intent{MaxCost}), seedable failure paths.
- 'How it compares': honest trade-off table vs LocalStack/moto/Real AWS that
  INCLUDES where Substrate is weaker (doesn't run workload code, by design) —
  framed as 'a different tier,' not better-than. Links to Scope & Philosophy.
- Renamed the differentiators heading 'Why Substrate' → 'Core capabilities'
  (the new subsection now owns the 'why').

Docs build clean (strict link check passes).
@scttfrdmn scttfrdmn merged commit f96ee6b into main Jun 4, 2026
9 checks passed
@scttfrdmn scttfrdmn deleted the docs/ai-fit-and-comparison branch June 4, 2026 00:18
@codecov
Copy link
Copy Markdown

codecov Bot commented Jun 4, 2026

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant