How to Use Kimi K2.7 Code in Instinct
Kimi K2.7 Code is built for debugging, refactoring, and implementing changes across multiple files. Compared with Kimi K2.6, it targets longer coding tasks; Moonshot recommends Kimi K2.6 for general writing and conversation.
Prompts to try with Kimi K2.7 Code in Instinct
Copy an example into Instinct, or ask in your own words.
Every morning, use Kimi K2.7 Code to review the latest test-run summaries. Group recurring failures, compare them with yesterday's findings, and propose the next diagnostic step.
Each weekday, use Kimi K2.7 Code to review newly reported software bugs. Extract reproduction details, identify missing evidence, and draft focused engineering tasks.
After each release, use Kimi K2.7 Code to compare incident summaries with the release's acceptance criteria. Identify behavior that needs a regression test.
Every week, use Kimi K2.7 Code to review dependency release notes supplied to this task. Identify changes that may affect our documented application behavior and propose migration checks.
Each evening, use Kimi K2.7 Code to revisit unresolved debugging hypotheses using today's incident updates. Mark each hypothesis as supported, contradicted, or still untested, with evidence.
How to set up Kimi K2.7 Code in Instinct
Pick the way that suits you. Each takes about two minutes.
Use the Spiral skill
- Send Instinct 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 Spiral's MCP server
- Create an API key
In the console, open API keys and create a key for Instinct. Copy it.
- Send Instinct this message
Instinct reads Spiral's skill and asks for your API key.
set up https://spiral.ai/SKILL.md
- Give Instinct the key
When Instinct asks for the key, add it where Instinct keeps secrets, never in the chat itself.
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 Kimi K2.7 Code
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": "openrouter", "model": "moonshotai/kimi-k2.7-code", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Instinct finds Kimi K2.7 Code on Spiral, shows the price and runs it.
Kimi K2.7 Code providers and prices
Spiral runs Kimi K2.7 Code through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.016002 per run at the default inputs. $0.78 per 1M input tokens, $3.90 per 1M output tokens.
Settings you can ask for
Ask Instinct for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Min p | 0 to 1 | ||
| Top k | a whole number from 0 | ||
| Top p | 0 to 1 | ||
| Logprobs | on or off | ||
| Max tokens | 1 to 235929 | The most output tokens to generate. | |
| Temperature | 0 to 2 | ||
| Top logprobs | 0 to 20 | ||
| Presence penalty | -2 to 2 | ||
| Frequency penalty | -2 to 2 | ||
| Repetition penalty | 0 to 2 | ||
| Parallel tool calls | on or off |
Every API field of MoonshotAI: Kimi K2.7 Code is on its tool page: open its card above.
What can you do with Kimi K2.7 Code in Instinct?
Kimi K2.7 Code can support recurring engineering investigations that continue across several runs. Give each run the latest evidence and a record of unresolved questions so it can propose the next useful action.
Fix a bug across multiple files
Ask Kimi K2.7 Code to trace a reported failure through the relevant code, propose a repair, and check it against the reproduction steps.
Implement a complete feature
Give Kimi K2.7 Code acceptance criteria for a feature that touches several components. Request the implementation and a report of which criteria were verified.
Refactor shared logic
Have Kimi K2.7 Code consolidate duplicated behavior while preserving existing interfaces and test expectations.
Investigate failing tests
Give Kimi K2.7 Code the failure output and relevant implementation. Ask for the likely cause, a focused correction, and regression coverage.
Plan a code migration
Ask Kimi K2.7 Code to identify affected components, order the changes, and define checks for each stage of a migration.
Review a proposed change
Ask Kimi K2.7 Code to compare a patch with its requirements and identify concrete behavior changes that need testing.
Investigate recurring failures
Use Kimi K2.7 Code to examine successive test or incident reports, track unresolved hypotheses, and propose the next diagnostic step.
More on Kimi K2.7 Code and Instinct
Overview
Kimi K2.7 Code in other agents
More Instinct guides
Questions about Kimi K2.7 Code in Instinct
Can Instinct use Kimi K2.7 Code?
Yes. Add Spiral to Instinct as an MCP server, and Instinct can find and run Kimi K2.7 Code like any built-in tool.
How much does Kimi K2.7 Code cost in Instinct?
Up to $0.018665 per run at the default inputs. $0.78 per 1M input tokens, $4.55 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Instinct plan is separate.
Do I need an OpenRouter account to use Kimi K2.7 Code in Instinct?
You need no OpenRouter account. Instinct uses one Spiral API key for every tool, and Spiral runs Kimi K2.7 Code on its own OpenRouter account.
How long does it take to set up Kimi K2.7 Code in Instinct?
About two minutes: create an API key, then send Instinct this message, then give Instinct the key.
Does Instinct wait while Kimi K2.7 Code runs?
Kimi K2.7 Code runs in the background. Instinct starts the run, checks on it and gives you the text, json when it's ready.
What is Kimi K2.7 Code best suited for?
Use it for coding tasks that require several connected changes, such as tracing a bug, refactoring a codebase, or implementing a feature across multiple files.
How much context can Kimi K2.7 Code handle?
It has a 256K-token context window. For code review, prioritize the relevant implementation, requirements, and test results so the task stays focused.
Can Kimi K2.7 Code run without thinking?
No. Kimi K2.7 Code always uses thinking mode.
Which sampling settings does Moonshot recommend?
Moonshot recommends temperature 1.0 and top_p 0.95. Use those as your starting point when evaluating coding results.
Does Kimi K2.7 Code retain reasoning between turns?
Yes. It preserves reasoning across turns, which supports coding tasks that continue through several rounds of investigation and changes.
Start using Kimi K2.7 Code in Instinct
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Instinct:
set up https://spiral.ai/SKILL.md


