You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello, just want to first thank the team for the wonderful job with vk.
Just wondering if the current set-up refers to any existing .claude directory for agent/subagent descriptions, and other agent 'configurations' (please correct me on the proper term to use) such as slash commands and agent skills. If so, can I check which directory does it refer to?
The reason for this is because I already have a several agent.md and skills.md that have been configured for specific tasks, and I was thinking if vk's orchestrator is able to use these description files, or do I need to re-define them in the 'Agent Profiles'.
My apologies if this was already mentioned somewhere in the docs or in another discussion!
Yes, VK references your existing .claude directories automatically — no need to re-define anything in Agent Profiles.
Here's how it works:
VK Scans Your .claude Directory
From the source code (slash_commands.rs), VK's discover_custom_command_descriptions() scans these locations:
Project-level: <workspace>/.claude/ → commands/*.md and skills/*/SKILL.md
Global: ~/.claude/ → same scan pattern
Plugins: each plugin's path and its .claude subdirectory
It extracts the YAML frontmatter description: from each .md file to populate the UI.
Claude Code Natively Reads .claude Too
VK spawns the official Claude Code CLI as a subprocess in your workspace directory, so Claude Code itself also picks up all your existing configs — CLAUDE.md, commands, skills, settings — from both project and global paths.
Agent Profiles Are a Separate Layer
Your .claude configs (slash commands, skills, project instructions) and VK's Agent Profiles (executor flags, permissions, model params) are complementary, not overlapping. You don't need to re-define your agents/skills in Agent Profiles.
Bottom Line
Your existing .claude/commands/*.md and .claude/skills/*/SKILL.md files work out of the box. Just keep them in your workspace (or ~/.claude/ for global use), and both VK and Claude Code will pick them up automatically.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
Hello, just want to first thank the team for the wonderful job with vk.
Just wondering if the current set-up refers to any existing
.claudedirectory for agent/subagent descriptions, and other agent 'configurations' (please correct me on the proper term to use) such as slash commands and agent skills. If so, can I check which directory does it refer to?The reason for this is because I already have a several agent.md and skills.md that have been configured for specific tasks, and I was thinking if vk's orchestrator is able to use these description files, or do I need to re-define them in the 'Agent Profiles'.
My apologies if this was already mentioned somewhere in the docs or in another discussion!
Thank you
All reactions