Prompt Chaining: Sequential LLM Pipelines with Validation Gates
A single unscoped prompt that tries to plan, draft, validate, and polish in one pass is a coin toss. Chain narrow calls instead, and put programmatic gates between them.
6 min read
blog.md · page 6 of 6
Writing by Subodh Jena on software engineering, LLMs, agent workflows, and the decisions behind building products.
A single unscoped prompt that tries to plan, draft, validate, and polish in one pass is a coin toss. Chain narrow calls instead, and put programmatic gates between them.
6 min read
The question is not which prompt to write. It is which tokens deserve a place in the context window on this turn. Context engineering is the discipline that answers that question.
7 min read
A free-text response that sometimes parses and sometimes does not is the single largest source of flakiness in agentic systems. Structured output makes the model behave like an ordinary function.
6 min read
A bare language model is a closed system. It cannot look up a fresh fact, cannot act in the world, cannot remember the last conversation. Every non-trivial agent is an attempt to lift one or more of those three constraints.
6 min read
The term agent has become overloaded, and the overloading is expensive. A system in which the language model controls its own flow costs more to run and more to debug. The default should be a workflow.
7 min read