Files
llmx/codex-rs/mcp-server/src/codex_tool_config.rs
Michael Bolin 0776d78357 feat: redesign sandbox config (#1373)
This is a major redesign of how sandbox configuration works and aims to
fix https://github.com/openai/codex/issues/1248. Specifically, it
replaces `sandbox_permissions` in `config.toml` (and the
`-s`/`--sandbox-permission` CLI flags) with a "table" with effectively
three variants:

```toml
# Safest option: full disk is read-only, but writes and network access are disallowed.
[sandbox]
mode = "read-only"

# The cwd of the Codex task is writable, as well as $TMPDIR on macOS.
# writable_roots can be used to specify additional writable folders.
[sandbox]
mode = "workspace-write"
writable_roots = []  # Optional, defaults to the empty list.
network_access = false  # Optional, defaults to false.

# Disable sandboxing: use at your own risk!!!
[sandbox]
mode = "danger-full-access"
```

This should make sandboxing easier to reason about. While we have
dropped support for `-s`, the way it works now is:

- no flags => `read-only`
- `--full-auto` => `workspace-write`
- currently, there is no way to specify `danger-full-access` via a CLI
flag, but we will revisit that as part of
https://github.com/openai/codex/issues/1254

Outstanding issue:

- As noted in the `TODO` on `SandboxPolicy::is_unrestricted()`, we are
still conflating sandbox preferences with approval preferences in that
case, which needs to be cleaned up.
2025-06-24 16:59:47 -07:00

204 lines
7.5 KiB
Rust

//! Configuration object accepted by the `codex` MCP tool-call.
use codex_core::protocol::AskForApproval;
use mcp_types::Tool;
use mcp_types::ToolInputSchema;
use schemars::JsonSchema;
use schemars::r#gen::SchemaSettings;
use serde::Deserialize;
use std::collections::HashMap;
use std::path::PathBuf;
use crate::json_to_toml::json_to_toml;
/// Client-supplied configuration for a `codex` tool-call.
#[derive(Debug, Clone, Deserialize, JsonSchema)]
#[serde(rename_all = "kebab-case")]
pub(crate) struct CodexToolCallParam {
/// The *initial user prompt* to start the Codex conversation.
pub prompt: String,
/// Optional override for the model name (e.g. "o3", "o4-mini").
#[serde(default, skip_serializing_if = "Option::is_none")]
pub model: Option<String>,
/// Configuration profile from config.toml to specify default options.
#[serde(default, skip_serializing_if = "Option::is_none")]
pub profile: Option<String>,
/// Working directory for the session. If relative, it is resolved against
/// the server process's current working directory.
#[serde(default, skip_serializing_if = "Option::is_none")]
pub cwd: Option<String>,
/// Execution approval policy expressed as the kebab-case variant name
/// (`unless-allow-listed`, `auto-edit`, `on-failure`, `never`).
#[serde(default, skip_serializing_if = "Option::is_none")]
pub approval_policy: Option<CodexToolCallApprovalPolicy>,
/// Individual config settings that will override what is in
/// CODEX_HOME/config.toml.
#[serde(default, skip_serializing_if = "Option::is_none")]
pub config: Option<HashMap<String, serde_json::Value>>,
}
// Custom enum mirroring `AskForApproval`, but constrained to the subset we
// expose via the tool-call schema.
#[derive(Debug, Clone, Deserialize, JsonSchema)]
#[serde(rename_all = "kebab-case")]
pub(crate) enum CodexToolCallApprovalPolicy {
AutoEdit,
UnlessAllowListed,
OnFailure,
Never,
}
impl From<CodexToolCallApprovalPolicy> for AskForApproval {
fn from(value: CodexToolCallApprovalPolicy) -> Self {
match value {
CodexToolCallApprovalPolicy::AutoEdit => AskForApproval::AutoEdit,
CodexToolCallApprovalPolicy::UnlessAllowListed => AskForApproval::UnlessAllowListed,
CodexToolCallApprovalPolicy::OnFailure => AskForApproval::OnFailure,
CodexToolCallApprovalPolicy::Never => AskForApproval::Never,
}
}
}
/// Builds a `Tool` definition (JSON schema etc.) for the Codex tool-call.
pub(crate) fn create_tool_for_codex_tool_call_param() -> Tool {
let schema = SchemaSettings::draft2019_09()
.with(|s| {
s.inline_subschemas = true;
s.option_add_null_type = false;
})
.into_generator()
.into_root_schema_for::<CodexToolCallParam>();
#[expect(clippy::expect_used)]
let schema_value =
serde_json::to_value(&schema).expect("Codex tool schema should serialise to JSON");
let tool_input_schema =
serde_json::from_value::<ToolInputSchema>(schema_value).unwrap_or_else(|e| {
panic!("failed to create Tool from schema: {e}");
});
Tool {
name: "codex".to_string(),
input_schema: tool_input_schema,
description: Some(
"Run a Codex session. Accepts configuration parameters matching the Codex Config struct.".to_string(),
),
annotations: None,
}
}
impl CodexToolCallParam {
/// Returns the initial user prompt to start the Codex conversation and the
/// effective Config object generated from the supplied parameters.
pub fn into_config(
self,
codex_linux_sandbox_exe: Option<PathBuf>,
) -> std::io::Result<(String, codex_core::config::Config)> {
let Self {
prompt,
model,
profile,
cwd,
approval_policy,
config: cli_overrides,
} = self;
// Build the `ConfigOverrides` recognised by codex-core.
let overrides = codex_core::config::ConfigOverrides {
model,
config_profile: profile,
cwd: cwd.map(PathBuf::from),
approval_policy: approval_policy.map(Into::into),
// Note we may want to expose a field on CodexToolCallParam to
// facilitate configuring the sandbox policy.
sandbox_policy: None,
model_provider: None,
codex_linux_sandbox_exe,
};
let cli_overrides = cli_overrides
.unwrap_or_default()
.into_iter()
.map(|(k, v)| (k, json_to_toml(v)))
.collect();
let cfg = codex_core::config::Config::load_with_cli_overrides(cli_overrides, overrides)?;
Ok((prompt, cfg))
}
}
#[cfg(test)]
mod tests {
use super::*;
use pretty_assertions::assert_eq;
/// We include a test to verify the exact JSON schema as "executable
/// documentation" for the schema. When can track changes to this test as a
/// way to audit changes to the generated schema.
///
/// Seeing the fully expanded schema makes it easier to casually verify that
/// the generated JSON for enum types such as "approval-policy" is compact.
/// Ideally, modelcontextprotocol/inspector would provide a simpler UI for
/// enum fields versus open string fields to take advantage of this.
///
/// As of 2025-05-04, there is an open PR for this:
/// https://github.com/modelcontextprotocol/inspector/pull/196
#[test]
fn verify_codex_tool_json_schema() {
let tool = create_tool_for_codex_tool_call_param();
#[expect(clippy::expect_used)]
let tool_json = serde_json::to_value(&tool).expect("tool serializes");
let expected_tool_json = serde_json::json!({
"name": "codex",
"description": "Run a Codex session. Accepts configuration parameters matching the Codex Config struct.",
"inputSchema": {
"type": "object",
"properties": {
"approval-policy": {
"description": "Execution approval policy expressed as the kebab-case variant name (`unless-allow-listed`, `auto-edit`, `on-failure`, `never`).",
"enum": [
"auto-edit",
"unless-allow-listed",
"on-failure",
"never"
],
"type": "string"
},
"config": {
"description": "Individual config settings that will override what is in CODEX_HOME/config.toml.",
"additionalProperties": true,
"type": "object"
},
"cwd": {
"description": "Working directory for the session. If relative, it is resolved against the server process's current working directory.",
"type": "string"
},
"model": {
"description": "Optional override for the model name (e.g. \"o3\", \"o4-mini\").",
"type": "string"
},
"profile": {
"description": "Configuration profile from config.toml to specify default options.",
"type": "string"
},
"prompt": {
"description": "The *initial user prompt* to start the Codex conversation.",
"type": "string"
},
},
"required": [
"prompt"
]
}
});
assert_eq!(expected_tool_json, tool_json);
}
}