Clinerules add tokens to every message, while workflows add them only when invoked. Knowing which to use keeps persistent context lean and turns repeatable multi-step processes into single slash commands.
Key takeaways · AI-distilled
A Cline workflow is injected into a single message inside <explicit_instructions> tags and then disappears, while clinerules append to the system promptThe standing instructions a model receives before any user input — defining its role, rules, tools, and tone for the whole conversation.Full definition →, so only clinerules cost tokenThe chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.Full definition → on every message.
Nick Baumann's rule of thumb: build a workflow when a task reads as 'first X, then Y, then Z', and keep clinerules for behavior that should shape every task, such as coding standards or TypeScript strict mode.
Global workflows live in ~/Documents/Cline/Workflows and workspace workflows in .clinerules/workflows/; the filename becomes the slash commandA named, reusable prompt invoked with a leading slash, turning a workflow you repeat into one word you type.Full definition →, and the workspace version wins when both share a name.
The example /pr-review.md workflow gathers the PR with gh pr view and gh pr diff, reads the original files for context windowThe maximum amount of text a model can consider at once — its working memory for the current conversation or task.Full definition →, analyzes the changes, then asks for confirmation before running gh pr review.
Terms in this piece · Glossary
system prompt — The standing instructions a model receives before any user input — defining its role, rules, tools, and tone for the whole conversation.
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
slash command — A named, reusable prompt invoked with a leading slash, turning a workflow you repeat into one word you type.
