Add "When not to build an agent" (Loop & Retry) to Blog

[loopandretry]

Adds one practitioner essay to the Blog section.

Link: https://loopandretry.github.io/posts/when-not-to-build-an-agent/

Why it fits: a decision framework for when a deterministic pipeline beats an autonomous agent — the same agent-architecture lane as the existing Lilian Weng and 8bitconcepts entries. Substantive full essay, formatted to match the existing non-GitHub Blog entries.

Disclosure: I write this blog; submitting because it is genuinely on-topic, not promotional.

[Jenqyang]

Decision: Not ready to merge yet.\n\nReason:\n1) The article is substantive and in scope: it provides concrete cost arithmetic, code comparisons, and a practical agent-versus-workflow checklist.\n2) The proposed URL includes a referral/tracking parameter (?ref=awesome-ai-agents), which should not be embedded in this curated list.\n\nNext step: Change the link to the canonical URL:\nhttps://loopandretry.github.io/posts/when-not-to-build-an-agent/

[loopandretry]

Thanks for the review — done. Swapped the link to the canonical URL (no tracking param): https://loopandretry.github.io/posts/when-not-to-build-an-agent/