Skip to content
Get packs

Guide · Agent skill

How to use the Kernel Analysis vphone600 agent skill

Updated · written with AI from the project's own documentation and checked against our catalogue data

The Kernel Analysis for vphone600 skill (skill name kernel-analysis-vphone600) gives an AI agent a structured way to study the vphone600 kernel reverse-engineering dataset in the Lakr233/vphone-cli repository. It is aimed at reverse engineers, researchers comparing kernel builds, and developers mapping raw addresses to known symbols. After this guide you will know what the skill does, how to install it, how to phrase requests so it triggers, and what your repository needs to contain.

Kernel Analysis for vphone600Lakr233/vphone-cli · MIT

Quick start

  1. 1Download and unzip the skill. Download the kernel-analysis-vphone600 ZIP from this page and unzip it into ~/.claude/skills/ or a project's .claude/skills/ folder.
  2. 2Check the repository dataset. Confirm Research/KernelSymbols contains kernel_symbols.db, kernel_index.tsv and the json folder with both vphone600 symbol files.
  3. 3Prepare the XNU reference. If Research/Reference/xnu is missing, run the shallow clone from GitHub, which is the only network step.
  4. 4Make a request. Ask about a vphone600 symbol or address and name the kernel, release or research, that you want used.
  5. 5Review the answer. Check that the kernel name, symbol and address are given, and that facts are separated from inferences.

What is an agent skill and what does this one do?

An agent skill is a folder containing a SKILL.md file with instructions that an AI agent reads when a request matches the skill's description. This skill is scoped to the vphone600 kernel artifacts stored in the repository.

Instead of guessing at function names, the agent first queries the local symbol database (Research/KernelSymbols/kernel_symbols.db) and the JSON symbol files for the release and research kernel builds. It then checks the matching behaviour in the open-source XNU source tree at Research/Reference/xnu. The skill also asks the agent to separate confirmed facts from its own inferences and to name the exact kernel it used.

  • Looks up function names and addresses in kernelcache.release.vphone600 and kernelcache.research.vphone600
  • Maps addresses to symbols and cross-references candidate code paths in XNU
  • Compares release and research builds for patch analysis
  • Labels findings as fact or inference

How do you install the skill?

Download the ZIP for kernel-analysis-vphone600 from this page and unzip it. For Claude Code, place the folder in ~/.claude/skills/ to make it available everywhere, or in a project's .claude/skills/ folder to limit it to that project.

In the Claude apps, you can instead upload the skill under Settings → Capabilities → Skills.

  • Personal install: unzip into ~/.claude/skills/
  • Project install: unzip into the project's .claude/skills/
  • Claude apps: Settings → Capabilities → Skills, then upload

What does your repository need?

The skill only works if the repository checkout contains the dataset it reads. Check these paths before you start.

If Research/Reference/xnu is missing, the skill's instructions tell the agent to create it with a shallow clone from GitHub. This is the only step that needs internet access.

  • Research/KernelSymbols/kernel_symbols.db
  • Research/KernelSymbols/kernel_index.tsv
  • Research/KernelSymbols/json/ with the release and research symbol JSON files
  • Research/Reference/xnu (or Git and internet access so it can be cloned)
  • SQLite or a tool that can query kernel_symbols.db
mkdir -p Research/Reference git clone --depth 1 https://github.com/apple-oss-distributions/xnu.git Research/Reference/xnu

How do you trigger the skill with natural requests?

Agents load a skill when your request matches its description. Mention the vphone600 kernel and the kind of task you want. Name the build if you care which one is used, and say so if you want a comparison.

The skill's own workflow confirms scope to vphone600 only, selects the release or research dataset by name, loads the linked JSON, and reports the kernel name, symbol and address.

  • "Using the vphone600 kernel symbols, what symbol is at this address in the release kernel?"
  • "Compare the release and research vphone600 kernelcache symbols for panic-related functions."
  • "Find the XNU source behaviour behind this vphone600 symbol and tell me which parts are inferred."
  • "Search the vphone600 research kernel JSON for symbols containing a given name."

What queries does the agent run?

The skill points to references/kernel-info-queries.md for reusable queries. These list the schema, count symbols, show matched and missed totals per kernel, and resolve a kernel name to its JSON file. They are meant to run from the repository root.

The agent uses the database to pick the dataset, then searches the JSON file. A typical fast search uses rg against the release symbol file.

  • Show table schema with .schema kernel_symbols
  • List kernel names with matched, missed, percent and total
  • Resolve the release or research JSON path by kernel name
  • Search symbols with rg -n against the JSON in Research/KernelSymbols/json/
sqlite3 Research/KernelSymbols/kernel_symbols.db "select json_path from kernel_symbols where kernel_name='kernelcache.release.vphone600';" rg -n 'panic' Research/KernelSymbols/json/kernelcache.release.vphone600.bin

What safety notes should you know?

This skill runs no scripts. It does use the network, but only when Research/Reference/xnu is missing and must be cloned from GitHub. Check the clone command before the agent runs it if you are working on a restricted network.

The skill supports analysis only. Its findings are partly inferred, so verify them independently before you change or ship any modified kernel. Do not rely on it for other devices or Apple platforms; it is limited to vphone600 by design.

  • No scripts are run by the skill
  • Network is used only for the optional XNU shallow clone
  • Analysis only; verify inferences before acting on them
  • Out of scope: other devices and Apple platforms

What tips help you get better results?

Name the kernel explicitly in your request, because the skill requires the output to say which build was used. Give exact addresses in hexadecimal when you have them.

The json_path values in the database and kernel_index.tsv may point to paths from the original machine. The skill resolves files under Research/KernelSymbols/json/ instead, so keep that folder intact in your checkout.

  • Name release or research in each request
  • Provide exact addresses and symbol names
  • Expect the answer to separate fact from inference
  • Keep the json folder structure unchanged

What people use it for

  • Mapping a raw address from a vphone600 kernel dump to its symbol name in the release build
  • Comparing symbols between the release and research kernelcache for patch analysis
  • Searching for panic-related functions in the vphone600 symbol JSON
  • Cross-checking a symbol's behaviour against the XNU source tree

Frequently asked questions

How do I install the kernel-analysis-vphone600 skill?

Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. In the Claude apps, you can upload it under Settings → Capabilities → Skills.

Is the kernel-analysis-vphone600 skill free?

The source material provided for this page does not state pricing or licence terms beyond the item's MIT licence. Check the repository's licence file for the terms that apply.

What does the skill require to work?

It needs a local checkout of the repository with Research/KernelSymbols, including kernel_symbols.db, kernel_index.tsv and the JSON symbol files. It also needs SQLite or a tool that can query the database, plus Git and internet access only if Research/Reference/xnu is missing.

Does the skill run scripts or use the network?

It runs no scripts. It uses the network only when Research/Reference/xnu is missing, in which case it performs a shallow clone from GitHub.

Can I use this skill for other devices or iOS kernels?

No. The skill is limited to vphone600 by design, and its instructions tell the agent to avoid claiming coverage outside vphone600 unless explicitly requested.

Why does the agent say the database paths are wrong?

The json_path values record absolute paths from when the symbols were generated, so they may not match your checkout. The skill resolves the JSON files under Research/KernelSymbols/json/ instead.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

This skill suits reverse engineers and researchers who already have the vphone600 Research/KernelSymbols dataset and want an agent to query it carefully and separate facts from inferences. It is a poor fit if you lack those files, need coverage of other devices or Apple platforms, or want an agent to produce kernels ready to ship without independent verification.

Download the skill