How to Use GPT-5 Mini in Factory Droid
GPT-5 Mini suits focused coding tasks, text extraction, and summaries with clear requirements. It responds faster than GPT-5 and works best when you specify exactly what a successful answer includes.
Prompts to try with GPT-5 Mini in Factory Droid
Copy an example into Factory Droid, or ask in your own words.
Use GPT-5 Mini to investigate the failing tests in this project. Identify the expected behavior, fix the underlying issue with a focused patch, and run the affected tests. Report the changes and actual test results.
Use GPT-5 Mini with high reasoning effort to trace why this project's retry logic can submit the same job twice. Add a regression test and a focused fix. Preserve the documented retry behavior.
Use GPT-5 Mini to write tests for the parser in this project. Cover empty input, malformed records, and missing required values using the existing test framework. Derive expected results from the parser's documented contract.
Use GPT-5 Mini to review the current diff for violations of the acceptance criteria in the issue description. Return only actionable findings, each with a file location, the violated requirement, and a suggested correction.
Use GPT-5 Mini to extract the public functions from this project's authentication module. Return JSON containing each function's name, parameters, return type, and documented errors. Mark undocumented behavior as unknown.
How to set up GPT-5 Mini in Factory Droid
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add the server
Droid saves it to your user config, ~/.factory/mcp.json, so it works in every project.
droid mcp add spiral https://api.spiral.ai/mcp --type http
- Sign in
Start droid, type /mcp, select spiral and authenticate it. Sign in to Spiral in the browser and click Authorize. No API key.
- Check it
Spiral shows as connected. needs authentication means the sign-in did not finish: run /mcp again.
droid mcp list
Use the Spiral skill
- Send Factory Droid this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run GPT-5 Mini
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "openai", "model": "openai/gpt-5-mini", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Factory Droid finds GPT-5 Mini on Spiral, shows the price and runs it.
GPT-5 Mini providers and prices
Spiral runs GPT-5 Mini through OpenAI. Open a card for today's price, with every fee included.
Up to $0.010662 per run at the default inputs. $0.325 per 1M input tokens, $0.0325 per 1M cached input tokens, $2.60 per 1M output tokens.
Settings you can ask for
Ask Factory Droid for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Max tokens | 16 to 128000 | The most output tokens to generate. | |
| Parallel tool calls | on or off |
Every API field of GPT-5 Mini is on its tool page: open its card above.
What can you do with GPT-5 Mini in Factory Droid?
Turn a reproducible bug or acceptance criterion into a focused patch and regression tests with GPT-5 Mini. Supply the relevant project context, define the expected behavior, and require evidence from checks.
Extract order details
Turn order messages into consistent records containing order numbers, requested changes, and dates. Require a missing value when the message does not provide a detail.
Sort support requests
Assign incoming messages to clearly defined categories such as billing, access, and product questions. Include a short supporting excerpt for each decision.
Summarize meeting decisions
Extract decisions, owners, deadlines, and unresolved questions from a transcript. Keep proposed actions separate from agreed commitments.
Draft answers from supplied policies
Prepare support reply drafts using the customer's message and current policy text. Flag questions the policy does not answer.
Fix a specific software bug
Investigate a reproducible failure, propose a focused patch, and check it against the expected behavior. Keep the task scoped to the reported issue.
Write regression tests
Turn a bug report and acceptance criteria into tests for the reported failure and relevant edge cases. Review the tests against the intended behavior.
Prepare a daily change digest
Compare supplied updates with the previous digest and report changes, unresolved items, and supporting excerpts. Use explicit rules to exclude repeated information.
More on GPT-5 Mini and Factory Droid
Overview
GPT-5 Mini in other agents
More Factory Droid guides
Questions about GPT-5 Mini in Factory Droid
Can Factory Droid use GPT-5 Mini?
Yes. Add Spiral to Factory Droid as an MCP server, and Factory Droid can find and run GPT-5 Mini like any built-in tool.
How much does GPT-5 Mini cost in Factory Droid?
Up to $0.010662 per run at the default inputs. $0.325 per 1M input tokens, $0.0325 per 1M cached input tokens, $2.60 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Factory Droid plan is separate.
Do I need an OpenAI account to use GPT-5 Mini in Factory Droid?
No. Spiral runs GPT-5 Mini on its own OpenAI account. You sign in to Spiral once and never paste a key.
How long does it take to set up GPT-5 Mini in Factory Droid?
About two minutes: add the server, then sign in, then check it.
Does Factory Droid wait while GPT-5 Mini runs?
GPT-5 Mini runs in the background. Factory Droid starts the run, checks on it and gives you the text, json when it's ready.
What is GPT-5 Mini best suited for?
Start with well-defined tasks: extract details from text, summarize a discussion against a checklist, or fix a specific coding issue. Give it the source material and a clear description of the result you need.
How much text can GPT-5 Mini handle?
Its context window is 400,000 tokens, and its maximum output is 128,000 tokens. Reasoning also consumes the generation budget, so that output limit does not mean 128,000 tokens of visible text.
Does GPT-5 Mini know about recent events?
Its built-in knowledge cutoff is May 31, 2024. Supply recent source material when dates, policies, or other changing details matter.
How should I prompt GPT-5 Mini?
State the task, separate the source text from your instructions, and specify the required format and length. Remove conflicting requirements and give a clear rule for missing information.
When should I give GPT-5 Mini more reasoning effort?
Test higher effort when a task involves difficult debugging or several dependent decisions. Lower effort is worth testing for straightforward work where less exploration is useful.
Start using GPT-5 Mini in Factory Droid
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Factory Droid:
set up https://spiral.ai/SKILL.md


