Codewhale documentation
ReleaseLatest release v0.10.1 · Oct 8, 2026These pages describe the v0.10.2 source candidate, which is not published yet.Changelog →

Move from Claude Code

Keep compatible project instructions and skill folders while choosing your own model connection. Review settings and plugin imports before applying them.

Keep your project instructions

AGENTS.md is the canonical project instruction file. CLAUDE.md and .claude/instructions.md are compatibility fallbacks; they do not override Codewhale security controls.

Plain .claude/skills/<name>/SKILL.md folders are discovered through the normal skills registry, with workspace trust and invocation controls.

Review a settings import

/import-claude

The command reads Claude settings and writes a reviewable report and unapplied config bundle. MCP servers use the existing import consent flow; permissions and hooks need separate review.

Note: Secret-shaped and unknown environment values are not imported or echoed. Running the plan does not enable servers or apply permissions.

Install the supported plugin subset

/plugin install <spec>

The declarative .claude-plugin/plugin.json subset supports skills, commands, agent profiles and MCP declarations. Codewhale review hashes and explicit trust and enablement still apply.

Note: Codewhale does not run Claude Code plugin runtimes. Hooks, plugin LSP declarations, build steps and TypeScript agents are outside this importer.

Read the compatibility limits. Choose credentials and endpoints in model setup.

Next

Manage skills

Inspect compatible directories before importing.

Connect a provider

Use your chosen hosted or local model.