Should this be an agent?
Six questions to choose a script, a workflow, an assistant or a bounded agent trial. Clear reasons, no hidden score.
Open tool ↗Local decision aid · no save
THE SOFT CAT WORKBENCH
Bring a question.
Make something useful.
Edit a prompt, check JSON, plan a context budget or inspect a cost estimate. Small tools with something you can take away.
7 tools need no API key. Chat Playground is the one that calls a model.
PICK THE JOB
8 tools. Each with a purpose.
No sign-up. Clear inputs. Inspectable outputs.
Chat Playground
Chat with models via OpenRouter using your own key. Compare responses and estimated costs.
Open tool ↗OpenRouter request · your key · provider charges
Prompt Workbench
Build reusable prompts, inspect filled messages and export an OpenRouter request. Save versions locally and keep a backup.
Open tool ↗Local editor · optional browser saves
Model Comparison
Compare recorded model prices against an editable workload. Filter by model ID or provider, then export the costs and sources.
Open tool ↗Local calculation · dated model rates
Context Budget Planner
Budget instructions, history, documents, tools and the reply. Compare the same request against dated context and output limits.
Open tool ↗Local plan · dated model limits
Token Cost Calculator
Estimate supplied text and repeated conversation history. Inspect parsed messages, distinguish missing replies and export the assumptions.
Open tool ↗Local text estimate · no save
Prompt Diff
Compare prompt words and whitespace, estimated token counts and signed input costs. Load saved Workbench versions and export the result.
Open tool ↗Local comparison · reads saved versions
JSON Output Validator
Check JSON output against draft-07 or 2020-12 schemas. Try broken examples and inspect exact constraint failures.
Open tool ↗Local schema check · no save
HOW THE PIECES WORK
Guides tied to the build.
How this workshop is published
The static site, browser tools, checked GitHub deployments and cloud maintenance without the retired server.
↗Planning a context budget
What the Context Budget Planner counts, how it handles conflicting limits and what its result can establish.
↗Reading a text-cost estimate
Follow repeated conversation history, inspect parsed messages and distinguish missing output from a free reply.
↗