Code Hunter by NightRaven
Test whether an exact library version works in your project before adopting it. Get an adopt, hold, or reject decision backed by a disposable trial and results for each requirement.
Free
Works with the AI Agents you already use
Code Hunter by NightRaven
Example session with this skill installed
I’m considering Zod 4.6.7 for a job PATCH validator currently using 3.25.76. Check whether the exact versions keep an explicitly supplied priority, strip UI-only metadata, and leave omitted priority absent so the API won’t overwrite a saved value. Use a disposable fixture.
- Read your context and instructions
- Compiled the code hunter by
Reject the existing PATCH schema on Zod 4.6.7. On a disposable Windows fixture using Node.js v26.13.0, I ran node dist/agensi-demo/code-hunter-zod-patch.cjs against the two installed package versions. Zod 3.25.76 passed all three checks. Zod 4.6.7 kept an explicit high priority and stripped uiOnly, but returned { id: 'J-1', priority: 'normal' } for { id: 'J-1' }. The expected result was { id: 'J-1' }.
The candidate with a default-free PATCH schema passed the same three checks. The original field was z.enum(['normal', 'high']).default('normal').optional(). The trial used z.enum(['normal', 'high']).optional() for PATCH and left any create-time default as a separate decision. This is consistent with Zod’s documented change to defaults.
Hold the migration until one real API test confirms that omitting priority preserves the stored value. The fixture exercised the schema in Node.js. It did not run the API or write to the project.
Connects securely to your tools. The creator never sees your data.
About this skill
Give Code Hunter a candidate library and a behavior your project needs. It reads the relevant project configuration, pins the library version, and turns each must-have into a check. With authorized execution, it runs the smallest representative trial in a disposable fixture or project copy.
You get an Adopt, Hold, or Reject decision for the tested version and configuration, with the exact commands, observed results, and gaps that still matter. It records upstream findings separately from the checks it ran. Your working project and lockfile stay untouched.
Code Hunter works with your agent’s existing tools and can compare candidates against the same assertions. If the agent lacks the right runtime or permission to execute, it prepares a reproduction and marks those checks Not run.
Once you’ve chosen what to adopt, my paid NightRaven skill helps your agent implement an approved change and record its verification. Browse my Agensi skills.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse All Claude Skills.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed today
- Free with an Agensi account
- Software
Node.js
Project runtime and test tools for an executable trial. Web or package registry access when public evidence or a package must be fetched.
Permissions
Reads authorized project context. Trial writes and commands stay in a disposable fixture or approved project copy. Public source research and package retrieval are optional.