AI assistant configuration reference

The Lola init container reads its configuration from environment variables. Set them on the init container in the eclipse-che custom resource, in the lola-config-default ConfigMap, or in a developer’s lola-config Secret.

Table 1. Variables
Variable Purpose Default

LOLA_DEFAULT_REPO_URL

URL of the recommended module repository

None

LOLA_DEFAULT_REPO_REF

Branch, tag, or commit of the recommended module. Without it, the container installs the default branch.

Default branch

LOLA_REPO_URL

Personal module repository of one developer, used instead of the recommended one

None

LOLA_REPO_REF

Branch, tag, or commit of the personal module. Without it, the container installs the default branch.

Default branch

LOLA_AGENTS

Assistants to install the module for: one name, a comma-separated list, or all

claude-code

LOLA_DISABLED

true skips the installation and leaves the installed content untouched. Any other value is ignored and logged.

None

LOLA_TIMEOUT

Timeout in seconds for each Lola command. 0 disables the timeout.

120

The object that supplies a variable decides which value wins when the same name is set twice:

  • An environment variable set on the init container in the eclipse-che custom resource wins over the same name in the ConfigMap or in the Secret. Kubernetes resolves an explicit env entry before the mounted ones.

  • LOLA_REPO_URL wins over LOLA_DEFAULT_REPO_URL: a developer points their own workspaces at a personal module without an administrator changing the recommended one, and deleting the key falls back to the recommended module.

  • With neither repository URL set, the container logs that it has nothing to do and exits without installing anything.

Choosing the AI assistants

The init container installs the module for the assistants that LOLA_AGENTS names: one assistant, a comma-separated subset, or all. Install the same module for every assistant rather than one module per assistant. With no value set, the container installs Claude Code.

Supported values are claude-code, opencode, gemini-cli, cursor, copilot-cli, copilot-vscode, openclaw, a comma-separated combination of them, and all.

  • all installs for every assistant that Lola installs by default, including assistants that Lola adds later.

  • all leaves out copilot-cli, because it writes to the same location as copilot-vscode. Add copilot-cli to the list explicitly to cover it.

  • An unknown name fails the installation: the status file reports failed and names the value.

The examples below show only the env part of the entry. Add it to the lola-init entry that you created in Install the recommended AI assistant configuration in every workspace, which keeps its image and its persistent-home mount.

A value set on the init container applies to every workspace in the instance. Leave LOLA_AGENTS unset there when developers must choose their own set.

One assistant for every workspace

Set LOLA_AGENTS on the init container.

spec:
  devEnvironments:
    initContainers:
      - name: lola-init
        env:
          - name: LOLA_AGENTS
            value: opencode
Several assistants in one workspace

List the assistants, separated by commas. Order and duplicates do not matter.

spec:
  devEnvironments:
    initContainers:
      - name: lola-init
        env:
          - name: LOLA_AGENTS
            value: claude-code,opencode
Every assistant that Lola supports

Set the value to all. The container then installs without naming a target, so one installation covers every assistant that Lola installs by default, except copilot-cli.

spec:
  devEnvironments:
    initContainers:
      - name: lola-init
        env:
          - name: LOLA_AGENTS
            value: all
A different set for some developers

Each developer adds LOLA_AGENTS to the lola-config Secret in their own namespace. The recommended module stays, and only the set of assistants changes. This works only while the init container does not set the variable itself. Adding LOLA_REPO_URL to the same Secret replaces the module as well. See Override the recommended AI assistant configuration.

stringData:
  LOLA_AGENTS: claude-code,gemini-cli

A change to the list behaves like a change of the repository or the ref: the init container removes the module from every assistant of the previous set, then installs it for the new one.

Destinations at user scope

The table lists the main destinations at user scope. The copilot-vscode target skips commands and MCP servers, and OpenClaw has no user scope: the init container installs it into /home/user/.openclaw/workspace, and it receives skills only.

Table 2. Assistant destinations at user scope
Assistant Destination in /home/user

claude-code

.claude/skills, .claude/commands, .claude/agents, a managed section of .claude/CLAUDE.md, and .mcp.json

opencode

.config/opencode/skills, .config/opencode/commands, .config/opencode/agents, a managed section of .config/opencode/AGENTS.md, and .config/opencode/opencode.json

gemini-cli

A managed section of GEMINI.md, .gemini/commands, and .gemini/settings.json

cursor

.cursor/skills, .cursor/commands, .cursor/agents, .cursor/rules, and .cursor/mcp.json

copilot-cli

.copilot/skills, .copilot/prompts, .copilot/agents, .copilot/copilot-instructions.md, and .copilot/mcp-config.json

copilot-vscode

.copilot/skills, .copilot/agents, and .copilot/copilot-instructions.md

openclaw

.openclaw/workspace/skills

Content that Lola does not carry

  • Assistant settings, such as the Claude Code settings.json file with hooks and permissions.

  • Plugins and hooks written as JavaScript or TypeScript files.

  • Credentials. Store them in a Secret, not in the module.

  • Absolute paths from your workstation. Lola copies Markdown text as written: it does not copy the referenced files and does not rewrite the paths. Use paths that exist in the workspace, such as /home/user.