llmx/codex-rs/core/src/openai_tools.rs

use serde::Deserialize;
use serde::Serialize;
use serde_json::Value as JsonValue;
use serde_json::json;
use std::collections::BTreeMap;
use std::collections::HashMap;

use crate::model_family::ModelFamily;
use crate::plan_tool::PLAN_TOOL;
use crate::protocol::AskForApproval;
use crate::protocol::SandboxPolicy;

#[derive(Debug, Clone, Serialize, PartialEq)]
pub struct ResponsesApiTool {
    pub(crate) name: String,
    pub(crate) description: String,
    /// TODO: Validation. When strict is set to true, the JSON schema,
    /// `required` and `additional_properties` must be present. All fields in
    /// `properties` must be present in `required`.
    pub(crate) strict: bool,
    pub(crate) parameters: JsonSchema,
}

/// When serialized as JSON, this produces a valid "Tool" in the OpenAI
/// Responses API.
#[derive(Debug, Clone, Serialize, PartialEq)]
#[serde(tag = "type")]
pub(crate) enum OpenAiTool {
    #[serde(rename = "function")]
    Function(ResponsesApiTool),
    #[serde(rename = "local_shell")]
    LocalShell {},
}

#[derive(Debug, Clone)]
pub enum ConfigShellToolType {
    DefaultShell,
    ShellWithRequest { sandbox_policy: SandboxPolicy },
    LocalShell,
}

#[derive(Debug, Clone)]
pub struct ToolsConfig {
    pub shell_type: ConfigShellToolType,
    pub plan_tool: bool,
}

impl ToolsConfig {
    pub fn new(
        model_family: &ModelFamily,
        approval_policy: AskForApproval,
        sandbox_policy: SandboxPolicy,
        include_plan_tool: bool,
    ) -> Self {
        let mut shell_type = if model_family.uses_local_shell_tool {
            ConfigShellToolType::LocalShell
        } else {
            ConfigShellToolType::DefaultShell
        };
        if matches!(approval_policy, AskForApproval::OnRequest) {
            shell_type = ConfigShellToolType::ShellWithRequest {
                sandbox_policy: sandbox_policy.clone(),
            }
        }

        Self {
            shell_type,
            plan_tool: include_plan_tool,
        }
    }
}

/// Generic JSON‑Schema subset needed for our tool definitions
#[derive(Debug, Clone, Serialize, Deserialize, PartialEq)]
#[serde(tag = "type", rename_all = "lowercase")]
pub(crate) enum JsonSchema {
    Boolean {
        #[serde(skip_serializing_if = "Option::is_none")]
        description: Option<String>,
    },
    String {
        #[serde(skip_serializing_if = "Option::is_none")]
        description: Option<String>,
    },
    /// MCP schema allows "number" | "integer" for Number
    #[serde(alias = "integer")]
    Number {
        #[serde(skip_serializing_if = "Option::is_none")]
        description: Option<String>,
    },
    Array {
        items: Box<JsonSchema>,

        #[serde(skip_serializing_if = "Option::is_none")]
        description: Option<String>,
    },
    Object {
        properties: BTreeMap<String, JsonSchema>,
        #[serde(skip_serializing_if = "Option::is_none")]
        required: Option<Vec<String>>,
        #[serde(
            rename = "additionalProperties",
            skip_serializing_if = "Option::is_none"
        )]
        additional_properties: Option<bool>,
    },
}

fn create_shell_tool() -> OpenAiTool {
    let mut properties = BTreeMap::new();
    properties.insert(
        "command".to_string(),
        JsonSchema::Array {
            items: Box::new(JsonSchema::String { description: None }),
            description: None,
        },
    );
    properties.insert(
        "workdir".to_string(),
        JsonSchema::String { description: None },
    );
    properties.insert(
        "timeout".to_string(),
        JsonSchema::Number { description: None },
    );

    OpenAiTool::Function(ResponsesApiTool {
        name: "shell".to_string(),
        description: "Runs a shell command and returns its output".to_string(),
        strict: false,
        parameters: JsonSchema::Object {
            properties,
            required: Some(vec!["command".to_string()]),
            additional_properties: Some(false),
        },
    })
}

fn create_shell_tool_for_sandbox(sandbox_policy: &SandboxPolicy) -> OpenAiTool {
    let mut properties = BTreeMap::new();
    properties.insert(
        "command".to_string(),
        JsonSchema::Array {
            items: Box::new(JsonSchema::String { description: None }),
            description: Some("The command to execute".to_string()),
        },
    );
    properties.insert(
        "workdir".to_string(),
        JsonSchema::String {
            description: Some("The working directory to execute the command in".to_string()),
        },
    );
    properties.insert(
        "timeout".to_string(),
        JsonSchema::Number {
            description: Some("The timeout for the command in milliseconds".to_string()),
        },
    );

    if matches!(sandbox_policy, SandboxPolicy::WorkspaceWrite { .. }) {
        properties.insert(
        "with_escalated_permissions".to_string(),
        JsonSchema::Boolean {
            description: Some("Whether to request escalated permissions. Set to true if command needs to be run without sandbox restrictions".to_string()),
        },
    );
        properties.insert(
        "justification".to_string(),
        JsonSchema::String {
            description: Some("Only set if ask_for_escalated_permissions is true. 1-sentence explanation of why we want to run this command.".to_string()),
        },
    );
    }

    let description = match sandbox_policy {
        SandboxPolicy::WorkspaceWrite {
            network_access,
            ..
        } => {
            format!(
                r#"
The shell tool is used to execute shell commands.
- When invoking the shell tool, your call will be running in a landlock sandbox, and some shell commands will require escalated privileges:
  - Types of actions that require escalated privileges:
    - Reading files outside the current directory
    - Writing files outside the current directory, and protected folders like .git or .env{}
  - Examples of commands that require escalated privileges:
    - git commit
    - npm install or pnpm install
    - cargo build
    - cargo test
- When invoking a command that will require escalated privileges:
  - Provide the with_escalated_permissions parameter with the boolean value true
  - Include a short, 1 sentence explanation for why we need to run with_escalated_permissions in the justification parameter."#,
                if !network_access {
                    "\n  - Commands that require network access\n"
                } else {
                    ""
                }
            )
        }
        SandboxPolicy::DangerFullAccess => {
            "Runs a shell command and returns its output.".to_string()
        }
        SandboxPolicy::ReadOnly => {
            r#"
The shell tool is used to execute shell commands.
- When invoking the shell tool, your call will be running in a landlock sandbox, and some shell commands (including apply_patch) will require escalated permissions:
  - Types of actions that require escalated privileges:
    - Reading files outside the current directory
    - Writing files
    - Applying patches
  - Examples of commands that require escalated privileges:
    - apply_patch
    - git commit
    - npm install or pnpm install
    - cargo build
    - cargo test
- When invoking a command that will require escalated privileges:
  - Provide the with_escalated_permissions parameter with the boolean value true
  - Include a short, 1 sentence explanation for why we need to run with_escalated_permissions in the justification parameter"#.to_string()
        }
    };

    OpenAiTool::Function(ResponsesApiTool {
        name: "shell".to_string(),
        description,
        strict: false,
        parameters: JsonSchema::Object {
            properties,
            required: Some(vec!["command".to_string()]),
            additional_properties: Some(false),
        },
    })
}

/// Returns JSON values that are compatible with Function Calling in the
/// Responses API:
/// https://platform.openai.com/docs/guides/function-calling?api-mode=responses
pub(crate) fn create_tools_json_for_responses_api(
    tools: &Vec<OpenAiTool>,
) -> crate::error::Result<Vec<serde_json::Value>> {
    let mut tools_json = Vec::new();

    for tool in tools {
        tools_json.push(serde_json::to_value(tool)?);
    }

    Ok(tools_json)
}

/// Returns JSON values that are compatible with Function Calling in the
/// Chat Completions API:
/// https://platform.openai.com/docs/guides/function-calling?api-mode=chat
pub(crate) fn create_tools_json_for_chat_completions_api(
    tools: &Vec<OpenAiTool>,
) -> crate::error::Result<Vec<serde_json::Value>> {
    // We start with the JSON for the Responses API and than rewrite it to match
    // the chat completions tool call format.
    let responses_api_tools_json = create_tools_json_for_responses_api(tools)?;
    let tools_json = responses_api_tools_json
        .into_iter()
        .filter_map(|mut tool| {
            if tool.get("type") != Some(&serde_json::Value::String("function".to_string())) {
                return None;
            }

            if let Some(map) = tool.as_object_mut() {
                // Remove "type" field as it is not needed in chat completions.
                map.remove("type");
                Some(json!({
                    "type": "function",
                    "function": map,
                }))
            } else {
                None
            }
        })
        .collect::<Vec<serde_json::Value>>();
    Ok(tools_json)
}

pub(crate) fn mcp_tool_to_openai_tool(
    fully_qualified_name: String,
    tool: mcp_types::Tool,
) -> Result<ResponsesApiTool, serde_json::Error> {
    let mcp_types::Tool {
        description,
        mut input_schema,
        ..
    } = tool;

    // OpenAI models mandate the "properties" field in the schema. The Agents
    // SDK fixed this by inserting an empty object for "properties" if it is not
    // already present https://github.com/openai/openai-agents-python/issues/449
    // so here we do the same.
    if input_schema.properties.is_none() {
        input_schema.properties = Some(serde_json::Value::Object(serde_json::Map::new()));
    }

    // Serialize to a raw JSON value so we can sanitize schemas coming from MCP
    // servers. Some servers omit the top-level or nested `type` in JSON
    // Schemas (e.g. using enum/anyOf), or use unsupported variants like
    // `integer`. Our internal JsonSchema is a small subset and requires
    // `type`, so we coerce/sanitize here for compatibility.
    let mut serialized_input_schema = serde_json::to_value(input_schema)?;
    sanitize_json_schema(&mut serialized_input_schema);
    let input_schema = serde_json::from_value::<JsonSchema>(serialized_input_schema)?;

    Ok(ResponsesApiTool {
        name: fully_qualified_name,
        description: description.unwrap_or_default(),
        strict: false,
        parameters: input_schema,
    })
}

/// Sanitize a JSON Schema (as serde_json::Value) so it can fit our limited
/// JsonSchema enum. This function:
/// - Ensures every schema object has a "type". If missing, infers it from
///   common keywords (properties => object, items => array, enum/const/format => string)
///   and otherwise defaults to "string".
/// - Fills required child fields (e.g. array items, object properties) with
///   permissive defaults when absent.
fn sanitize_json_schema(value: &mut JsonValue) {
    match value {
        JsonValue::Bool(_) => {
            // JSON Schema boolean form: true/false. Coerce to an accept-all string.
            *value = json!({ "type": "string" });
        }
        JsonValue::Array(arr) => {
            for v in arr.iter_mut() {
                sanitize_json_schema(v);
            }
        }
        JsonValue::Object(map) => {
            // First, recursively sanitize known nested schema holders
            if let Some(props) = map.get_mut("properties") {
                if let Some(props_map) = props.as_object_mut() {
                    for (_k, v) in props_map.iter_mut() {
                        sanitize_json_schema(v);
                    }
                }
            }
            if let Some(items) = map.get_mut("items") {
                sanitize_json_schema(items);
            }
            // Some schemas use oneOf/anyOf/allOf - sanitize their entries
            for combiner in ["oneOf", "anyOf", "allOf", "prefixItems"] {
                if let Some(v) = map.get_mut(combiner) {
                    sanitize_json_schema(v);
                }
            }

            // Normalize/ensure type
            let mut ty = map
                .get("type")
                .and_then(|v| v.as_str())
                .map(|s| s.to_string());

            // If type is an array (union), pick first supported; else leave to inference
            if ty.is_none() {
                if let Some(JsonValue::Array(types)) = map.get("type") {
                    for t in types {
                        if let Some(tt) = t.as_str() {
                            if matches!(
                                tt,
                                "object" | "array" | "string" | "number" | "integer" | "boolean"
                            ) {
                                ty = Some(tt.to_string());
                                break;
                            }
                        }
                    }
                }
            }

            // Infer type if still missing
            if ty.is_none() {
                if map.contains_key("properties")
                    || map.contains_key("required")
                    || map.contains_key("additionalProperties")
                {
                    ty = Some("object".to_string());
                } else if map.contains_key("items") || map.contains_key("prefixItems") {
                    ty = Some("array".to_string());
                } else if map.contains_key("enum")
                    || map.contains_key("const")
                    || map.contains_key("format")
                {
                    ty = Some("string".to_string());
                } else if map.contains_key("minimum")
                    || map.contains_key("maximum")
                    || map.contains_key("exclusiveMinimum")
                    || map.contains_key("exclusiveMaximum")
                    || map.contains_key("multipleOf")
                {
                    ty = Some("number".to_string());
                }
            }
            // If we still couldn't infer, default to string
            let ty = ty.unwrap_or_else(|| "string".to_string());
            map.insert("type".to_string(), JsonValue::String(ty.to_string()));

            // Ensure object schemas have properties map
            if ty == "object" {
                if !map.contains_key("properties") {
                    map.insert(
                        "properties".to_string(),
                        JsonValue::Object(serde_json::Map::new()),
                    );
                }
                // If additionalProperties is an object schema, sanitize it too.
                // Leave booleans as-is, since JSON Schema allows boolean here.
                if let Some(ap) = map.get_mut("additionalProperties") {
                    let is_bool = matches!(ap, JsonValue::Bool(_));
                    if !is_bool {
                        sanitize_json_schema(ap);
                    }
                }
            }

            // Ensure array schemas have items
            if ty == "array" && !map.contains_key("items") {
                map.insert("items".to_string(), json!({ "type": "string" }));
            }
        }
        _ => {}
    }
}

/// Returns a list of OpenAiTools based on the provided config and MCP tools.
/// Note that the keys of mcp_tools should be fully qualified names. See
/// [`McpConnectionManager`] for more details.
pub(crate) fn get_openai_tools(
    config: &ToolsConfig,
    mcp_tools: Option<HashMap<String, mcp_types::Tool>>,
) -> Vec<OpenAiTool> {
    let mut tools: Vec<OpenAiTool> = Vec::new();

    match &config.shell_type {
        ConfigShellToolType::DefaultShell => {
            tools.push(create_shell_tool());
        }
        ConfigShellToolType::ShellWithRequest { sandbox_policy } => {
            tools.push(create_shell_tool_for_sandbox(sandbox_policy));
        }
        ConfigShellToolType::LocalShell => {
            tools.push(OpenAiTool::LocalShell {});
        }
    }

    if config.plan_tool {
        tools.push(PLAN_TOOL.clone());
    }

    if let Some(mcp_tools) = mcp_tools {
        for (name, tool) in mcp_tools {
            match mcp_tool_to_openai_tool(name.clone(), tool.clone()) {
                Ok(converted_tool) => tools.push(OpenAiTool::Function(converted_tool)),
                Err(e) => {
                    tracing::error!("Failed to convert {name:?} MCP tool to OpenAI tool: {e:?}");
                }
            }
        }
    }

    tools
}

#[cfg(test)]
#[allow(clippy::expect_used)]
mod tests {
    use crate::model_family::find_family_for_model;
    use mcp_types::ToolInputSchema;
    use pretty_assertions::assert_eq;

    use super::*;

    fn assert_eq_tool_names(tools: &[OpenAiTool], expected_names: &[&str]) {
        let tool_names = tools
            .iter()
            .map(|tool| match tool {
                OpenAiTool::Function(ResponsesApiTool { name, .. }) => name,
                OpenAiTool::LocalShell {} => "local_shell",
            })
            .collect::<Vec<_>>();

        assert_eq!(
            tool_names.len(),
            expected_names.len(),
            "tool_name mismatch, {tool_names:?}, {expected_names:?}",
        );
        for (name, expected_name) in tool_names.iter().zip(expected_names.iter()) {
            assert_eq!(
                name, expected_name,
                "tool_name mismatch, {name:?}, {expected_name:?}"
            );
        }
    }

    #[test]
    fn test_get_openai_tools() {
        let model_family = find_family_for_model("codex-mini-latest")
            .expect("codex-mini-latest should be a valid model family");
        let config = ToolsConfig::new(
            &model_family,
            AskForApproval::Never,
            SandboxPolicy::ReadOnly,
            true,
        );
        let tools = get_openai_tools(&config, Some(HashMap::new()));

        assert_eq_tool_names(&tools, &["local_shell", "update_plan"]);
    }

    #[test]
    fn test_get_openai_tools_default_shell() {
        let model_family = find_family_for_model("o3").expect("o3 should be a valid model family");
        let config = ToolsConfig::new(
            &model_family,
            AskForApproval::Never,
            SandboxPolicy::ReadOnly,
            true,
        );
        let tools = get_openai_tools(&config, Some(HashMap::new()));

        assert_eq_tool_names(&tools, &["shell", "update_plan"]);
    }

    #[test]
    fn test_get_openai_tools_mcp_tools() {
        let model_family = find_family_for_model("o3").expect("o3 should be a valid model family");
        let config = ToolsConfig::new(
            &model_family,
            AskForApproval::Never,
            SandboxPolicy::ReadOnly,
            false,
        );
        let tools = get_openai_tools(
            &config,
            Some(HashMap::from([(
                "test_server/do_something_cool".to_string(),
                mcp_types::Tool {
                    name: "do_something_cool".to_string(),
                    input_schema: ToolInputSchema {
                        properties: Some(serde_json::json!({
                            "string_argument": {
                                "type": "string",
                            },
                            "number_argument": {
                                "type": "number",
                            },
                            "object_argument": {
                                "type": "object",
                                "properties": {
                                    "string_property": { "type": "string" },
                                    "number_property": { "type": "number" },
                                },
                                "required": [
                                    "string_property",
                                    "number_property"
                                ],
                                "additionalProperties": Some(false),
                            },
                        })),
                        required: None,
                        r#type: "object".to_string(),
                    },
                    output_schema: None,
                    title: None,
                    annotations: None,
                    description: Some("Do something cool".to_string()),
                },
            )])),
        );

        assert_eq_tool_names(&tools, &["shell", "test_server/do_something_cool"]);

        assert_eq!(
            tools[1],
            OpenAiTool::Function(ResponsesApiTool {
                name: "test_server/do_something_cool".to_string(),
                parameters: JsonSchema::Object {
                    properties: BTreeMap::from([
                        (
                            "string_argument".to_string(),
                            JsonSchema::String { description: None }
                        ),
                        (
                            "number_argument".to_string(),
                            JsonSchema::Number { description: None }
                        ),
                        (
                            "object_argument".to_string(),
                            JsonSchema::Object {
                                properties: BTreeMap::from([
                                    (
                                        "string_property".to_string(),
                                        JsonSchema::String { description: None }
                                    ),
                                    (
                                        "number_property".to_string(),
                                        JsonSchema::Number { description: None }
                                    ),
                                ]),
                                required: Some(vec![
                                    "string_property".to_string(),
                                    "number_property".to_string(),
                                ]),
                                additional_properties: Some(false),
                            },
                        ),
                    ]),
                    required: None,
                    additional_properties: None,
                },
                description: "Do something cool".to_string(),
                strict: false,
            })
        );
    }

    #[test]
    fn test_mcp_tool_property_missing_type_defaults_to_string() {
        let model_family = find_family_for_model("o3").expect("o3 should be a valid model family");
        let config = ToolsConfig::new(
            &model_family,
            AskForApproval::Never,
            SandboxPolicy::ReadOnly,
            false,
        );

        let tools = get_openai_tools(
            &config,
            Some(HashMap::from([(
                "dash/search".to_string(),
                mcp_types::Tool {
                    name: "search".to_string(),
                    input_schema: ToolInputSchema {
                        properties: Some(serde_json::json!({
                            "query": {
                                "description": "search query"
                            }
                        })),
                        required: None,
                        r#type: "object".to_string(),
                    },
                    output_schema: None,
                    title: None,
                    annotations: None,
                    description: Some("Search docs".to_string()),
                },
            )])),
        );

        assert_eq_tool_names(&tools, &["shell", "dash/search"]);

        assert_eq!(
            tools[1],
            OpenAiTool::Function(ResponsesApiTool {
                name: "dash/search".to_string(),
                parameters: JsonSchema::Object {
                    properties: BTreeMap::from([(
                        "query".to_string(),
                        JsonSchema::String {
                            description: Some("search query".to_string())
                        }
                    )]),
                    required: None,
                    additional_properties: None,
                },
                description: "Search docs".to_string(),
                strict: false,
            })
        );
    }

    #[test]
    fn test_mcp_tool_integer_normalized_to_number() {
        let model_family = find_family_for_model("o3").expect("o3 should be a valid model family");
        let config = ToolsConfig::new(
            &model_family,
            AskForApproval::Never,
            SandboxPolicy::ReadOnly,
            false,
        );

        let tools = get_openai_tools(
            &config,
            Some(HashMap::from([(
                "dash/paginate".to_string(),
                mcp_types::Tool {
                    name: "paginate".to_string(),
                    input_schema: ToolInputSchema {
                        properties: Some(serde_json::json!({
                            "page": { "type": "integer" }
                        })),
                        required: None,
                        r#type: "object".to_string(),
                    },
                    output_schema: None,
                    title: None,
                    annotations: None,
                    description: Some("Pagination".to_string()),
                },
            )])),
        );

        assert_eq_tool_names(&tools, &["shell", "dash/paginate"]);
        assert_eq!(
            tools[1],
            OpenAiTool::Function(ResponsesApiTool {
                name: "dash/paginate".to_string(),
                parameters: JsonSchema::Object {
                    properties: BTreeMap::from([(
                        "page".to_string(),
                        JsonSchema::Number { description: None }
                    )]),
                    required: None,
                    additional_properties: None,
                },
                description: "Pagination".to_string(),
                strict: false,
            })
        );
    }

    #[test]
    fn test_mcp_tool_array_without_items_gets_default_string_items() {
        let model_family = find_family_for_model("o3").expect("o3 should be a valid model family");
        let config = ToolsConfig::new(
            &model_family,
            AskForApproval::Never,
            SandboxPolicy::ReadOnly,
            false,
        );

        let tools = get_openai_tools(
            &config,
            Some(HashMap::from([(
                "dash/tags".to_string(),
                mcp_types::Tool {
                    name: "tags".to_string(),
                    input_schema: ToolInputSchema {
                        properties: Some(serde_json::json!({
                            "tags": { "type": "array" }
                        })),
                        required: None,
                        r#type: "object".to_string(),
                    },
                    output_schema: None,
                    title: None,
                    annotations: None,
                    description: Some("Tags".to_string()),
                },
            )])),
        );

        assert_eq_tool_names(&tools, &["shell", "dash/tags"]);
        assert_eq!(
            tools[1],
            OpenAiTool::Function(ResponsesApiTool {
                name: "dash/tags".to_string(),
                parameters: JsonSchema::Object {
                    properties: BTreeMap::from([(
                        "tags".to_string(),
                        JsonSchema::Array {
                            items: Box::new(JsonSchema::String { description: None }),
                            description: None
                        }
                    )]),
                    required: None,
                    additional_properties: None,
                },
                description: "Tags".to_string(),
                strict: false,
            })
        );
    }

    #[test]
    fn test_mcp_tool_anyof_defaults_to_string() {
        let model_family = find_family_for_model("o3").expect("o3 should be a valid model family");
        let config = ToolsConfig::new(
            &model_family,
            AskForApproval::Never,
            SandboxPolicy::ReadOnly,
            false,
        );

        let tools = get_openai_tools(
            &config,
            Some(HashMap::from([(
                "dash/value".to_string(),
                mcp_types::Tool {
                    name: "value".to_string(),
                    input_schema: ToolInputSchema {
                        properties: Some(serde_json::json!({
                            "value": { "anyOf": [ { "type": "string" }, { "type": "number" } ] }
                        })),
                        required: None,
                        r#type: "object".to_string(),
                    },
                    output_schema: None,
                    title: None,
                    annotations: None,
                    description: Some("AnyOf Value".to_string()),
                },
            )])),
        );

        assert_eq_tool_names(&tools, &["shell", "dash/value"]);
        assert_eq!(
            tools[1],
            OpenAiTool::Function(ResponsesApiTool {
                name: "dash/value".to_string(),
                parameters: JsonSchema::Object {
                    properties: BTreeMap::from([(
                        "value".to_string(),
                        JsonSchema::String { description: None }
                    )]),
                    required: None,
                    additional_properties: None,
                },
                description: "AnyOf Value".to_string(),
                strict: false,
            })
        );
    }
}