Workflows
Split work into linked prompts only when a step needs checking, and build a SaaS one working slice at a time.
Build a SaaS
Write a spec, design the database, build one user action at a time, and use established providers for login and billing.
4 techniques →-
How Do You Build a SaaS With Claude, One Slice at a Time? Build one user action at a time, end to end, from database to screen, and test it before starting the next. Level 4: Multi-tool systemsBy layernothing runs until the end
- all the tables
- all the endpoints
- all the screens
By slicesomething works each day- one user action
- database to screen
- tested, then committed
-
How Do You Design a SaaS Database With Claude? Before any screens, have Claude draft the schema from your spec, including who owns each row. Level 4: Multi-tool systemsCheap to change laterCostly to change later
- A screen
- An empty table
- A table with live data
-
Should Claude Write Your SaaS Login and Billing Code? Claude shouldn't write them from scratch. Level 4: Multi-tool systemsWritten from scratchyour liability
- password hashing
- session handling
- card and retry logic
Provider plus Claudereviewable- hosted auth and checkout
- webhook signature checks
- tests for failure paths
-
How Do You Write a SaaS Spec With Claude Before Coding? Have Claude interview you until it can write a one-page spec to SPEC.md, and cut that spec down yourself. Level 4: Multi-tool systems
- 1Claude interviews you
- 2It writes SPEC.md
- 3Build in a fresh session
- 4Review the diff against SPEC.md
Prompt chaining
Know when splitting a task helps, pass only the fields the next step needs, add quality checks, and run parts in parallel.
4 techniques →-
How Do You Add a Quality Check Between Claude Prompts? Add a step that tests the output against stated criteria before anything moves on. Level 3: Workflows
- 1Generate a draft
- 2Review against criteria
- 3Refine from the review
Fix once or twice, then send to a person -
How Do You Run Claude Prompts in Parallel and Merge Them? Send each independent part to its own prompt, then combine the results in code or with one final merge prompt. Level 3: Workflows
- 50%lower cost than standard calls
- 100,000requests per batchor 256 MB, whichever is first
- 1 hourmost batches finish within
- 24 hoursthen unfinished work expires
-
How Do You Pass Results From One Claude Prompt to the Next? Agree on the fields that the next step needs, have Claude output exactly those, and pass only that to the next prompt. Level 3: WorkflowsProse summaryhope nothing was lost
- A missing point stays hidden
- Abandoned ideas come along
- More tokens at each step
Agreed fieldsonly what the next step needs- An empty field shows the gap
- The next step starts clean
- Fewer tokens at each step
-
When Should You Split a Claude Task Into Separate Prompts? Split a task only when you need to see or check the output in the middle, or when the pipeline has to run the same way every time. Level 3: Workflows
- The output changes typequotes become a report
- A person should approvespot-check before moving on
- A cheaper model can take overone step needs less power
Can you build a real workflow with Claude?
Six questions on building a SaaS with Claude and on when chaining prompts is worth it. Each answer comes right after you pick, with no account needed.
When a workflow is worth building
A workflow makes sense when you run the same job repeatedly, or when one job has stages you need to check. Anthropic now says Claude handles most multi-step reasoning on its own, so try a single clear prompt first. Move to a workflow when you can name the point where that prompt fails, or when you need to log and inspect what happens in the middle.
Two lessons, two scales
Prompt chaining covers a job that takes minutes: how to split it, pass results along, add a check, and run parts in parallel. Building a SaaS covers a job that takes weeks, mostly in Claude Code: a written spec, a reviewed database design, one working feature at a time, and login and billing handled by established providers.
What is a Claude workflow?
It's a repeatable way of using Claude for a job with several stages, where each stage has a known input and output. It can be a set of prompts you run by hand, a script calling the API, or a Claude Code session working from a written spec.
Is prompt chaining still necessary with Claude?
Less often than it was. Anthropic says adaptive thinking and subagents handle most multi-step reasoning, and that explicit chaining is still useful when you need to inspect intermediate outputs or enforce a fixed pipeline. Try one prompt first.
Can Claude build a whole application?
Claude Code can write most of an application, but it works best from a written spec, one feature at a time, with tests it can run. You still need to review the database design, security and anything that touches money.
Which Claude model should I use for a workflow?
Match each step to its difficulty. Anthropic describes Haiku for high-volume classification and extraction, Sonnet for a balance of speed and intelligence, and Opus for long-running work. Test on your own task, since many steps show no visible difference between models.