conductor-oss/conductor · error · IllegalArgumentException

Model string cannot be null or empty

Error message

Model string cannot be null or empty

What it means

ModelParser.parse rejects a null or blank model string outright. The parser expects a 'provider/model' formatted identifier and fails fast before any splitting when the input is absent or whitespace-only. Mirrors the Python model_parser.

Source

Thrown at common/src/main/java/org/conductoross/conductor/common/metadata/agent/ModelParser.java:40

public class ModelParser {

    @Data
    @AllArgsConstructor
    public static class ParsedModel {
        private String provider;
        private String model;
    }

    /**
     * Parse a model string like "openai/gpt-4o" into provider and model.
     *
     * @param modelString The model string in "provider/model" format.
     * @return A ParsedModel with provider and model components.
     * @throws IllegalArgumentException if the format is invalid.
     */
    public static ParsedModel parse(String modelString) {
        if (modelString == null || modelString.isBlank()) {
            throw new IllegalArgumentException("Model string cannot be null or empty");
        }

        int slashIdx = modelString.indexOf('/');
        if (slashIdx <= 0 || slashIdx >= modelString.length() - 1) {
            throw new IllegalArgumentException(
                    "Invalid model format: '" + modelString + "'. Expected 'provider/model'.");
        }

        String provider = modelString.substring(0, slashIdx);
        String model = modelString.substring(slashIdx + 1);
        return new ParsedModel(provider, model);
    }
}

View on GitHub (pinned to cf7c3e4a8a)

Solutions

1. Provide a non-empty model string in 'provider/model' form (e.g. 'openai/gpt-4o').
2. Validate/require the model config field at agent-definition load time.
3. Fall back to a documented default model when the configured value is absent.

Example fix

// before: null/blank model throws
ParsedModel m = ModelParser.parse(agent.getModel());

// after: guard then default
String model = agent.getModel();
if (model == null || model.isBlank()) {
    model = DEFAULT_MODEL; // e.g. "openai/gpt-4o"
}
ParsedModel m = ModelParser.parse(model);

// Guard against null/blank before parsing
static boolean isNonBlankModel(String s) {
    return s != null && !s.isBlank();
}
if (!isNonBlankModel(modelString)) {
    throw new IllegalArgumentException("model must be a non-empty 'provider/model' string");
}

// Narrow to a present, non-blank model string
static boolean isParsableModelString(String s) {
    return s != null && !s.isBlank()
        && s.indexOf('/') > 0 && s.indexOf('/') < s.length() - 1;
}

// Catch missing model and apply a default
try {
    return ModelParser.parse(modelString);
} catch (IllegalArgumentException e) {
    return ModelParser.parse(DEFAULT_MODEL);
}

When it happens

Trigger: Calling ModelParser.parse(null) or ModelParser.parse("") / parse(" ") from agent/AI-task configuration resolution.

Common situations: An agent task definition with a missing or empty 'model' field; a config default that resolves to null; environment variable for the model not set.

Related errors

• Invalid model format: '%s'. Expected 'provider/model'.
• skillRef is required
• workflowDefinition is a string, but not a valid DSL string
• workflowDefinition must be either null, or WorkflowDef, or a
• priority MUST be between 0 and 99 (inclusive)

AI-assisted analysis of conductor-oss/conductor@cf7c3e4a8a (2026-08-14). Data as JSON: /api/errors/5ccec2de7b9b8bde. Report an issue: GitHub.