# Review request: is this a fair LangGraph setup for a "blog post to social posts" benchmark?

**URL:** <https://forum.langchain.com/t/review-request-is-this-a-fair-langgraph-setup-for-a-blog-post-to-social-posts-benchmark/4847>\
**Category:** LangGraph\
**Tags:** python-help\
**Created:** [October 27, 2026, 3:12pm UTC](https://forum.langchain.com/t/review-request-is-this-a-fair-langgraph-setup-for-a-blog-post-to-social-posts-benchmark/4847 "2026-10-27T15:12:18Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![kelvindesman](https://yyz1.discourse-cdn.com/flex007/user_avatar/forum.langchain.com/kelvindesman/32/3711_2.png) [@kelvindesman](https://forum.langchain.com/u/kelvindesman)\
**Post date:** [October 27, 2026, 3:12pm UTC](https://forum.langchain.com/t/review-request-is-this-a-fair-langgraph-setup-for-a-blog-post-to-social-posts-benchmark/4847/1 "2026-10-27T15:12:18Z")

</div>

Hi all. We’re comparing four ways to turn one blog post into a LinkedIn post, an X thread and a 5-slide carousel: a plain cron job, cron plus one LLM call, a LangGraph graph, and an autonomous agent (Hermes Agent). Same models, same metering proxy, same gates for every arm. The design was pre-registered before any counted run.

Our LangGraph arm is a StateGraph: fetch, draft, validate (code), critique (LLM), revise, with at most 2 revisions and `recursion_limit` 20. It uses `ChatOpenAI` against an OpenAI-compatible Workers AI endpoint (llama-4-scout and deepseek-v4-flash).

- Graph code (about 100 lines): [social-publish-4way-2026-10/arms/arm3\_langgraph.py at main · kelvindesman/social-publish-4way-2026-10 · GitHub](https://github.com/kelvindesman/social-publish-4way-2026-10/blob/main/arms/arm3_langgraph.py)
- Pre-registration and deviations: [social-publish-4way-2026-10/PREREGISTRATION.md at main · kelvindesman/social-publish-4way-2026-10 · GitHub](https://github.com/kelvindesman/social-publish-4way-2026-10/blob/main/PREREGISTRATION.md)
- Per-run results: [https://lokerdollar.com/data/social-publish-4way-2026-10.csv](https://lokerdollar.com/data/social-publish-4way-2026-10.csv)

Two questions:

1. Is there anything in the graph that a LangGraph user would call a strawman, such as a node or edge you’d build differently for this job?
2. Is anything missing that would change publishability or cost, like structured output or a different retry policy?

We’ll quote or link your answer in the write-up, including disagreement. Thanks for looking.