Files
codex/codex-rs/core/src/tools/context.rs
pakrym-oai d71e042694 Enforce single tool output type in codex handlers (#14157)
We'll need to associate output schema with each tool. Each tool can only
have on output type.
2026-03-09 21:49:44 -07:00

340 lines
10 KiB
Rust

use crate::codex::Session;
use crate::codex::TurnContext;
use crate::tools::TELEMETRY_PREVIEW_MAX_BYTES;
use crate::tools::TELEMETRY_PREVIEW_MAX_LINES;
use crate::tools::TELEMETRY_PREVIEW_TRUNCATION_NOTICE;
use crate::turn_diff_tracker::TurnDiffTracker;
use codex_protocol::mcp::CallToolResult;
use codex_protocol::models::FunctionCallOutputBody;
use codex_protocol::models::FunctionCallOutputContentItem;
use codex_protocol::models::FunctionCallOutputPayload;
use codex_protocol::models::ResponseInputItem;
use codex_protocol::models::ShellToolCallParams;
use codex_protocol::models::function_call_output_content_items_to_text;
use codex_utils_string::take_bytes_at_char_boundary;
use std::borrow::Cow;
use std::sync::Arc;
use tokio::sync::Mutex;
pub type SharedTurnDiffTracker = Arc<Mutex<TurnDiffTracker>>;
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
pub enum ToolCallSource {
Direct,
JsRepl,
CodeMode,
}
#[derive(Clone)]
pub struct ToolInvocation {
pub session: Arc<Session>,
pub turn: Arc<TurnContext>,
pub tracker: SharedTurnDiffTracker,
pub call_id: String,
pub tool_name: String,
pub payload: ToolPayload,
}
#[derive(Clone, Debug)]
pub enum ToolPayload {
Function {
arguments: String,
},
Custom {
input: String,
},
LocalShell {
params: ShellToolCallParams,
},
Mcp {
server: String,
tool: String,
raw_arguments: String,
},
}
impl ToolPayload {
pub fn log_payload(&self) -> Cow<'_, str> {
match self {
ToolPayload::Function { arguments } => Cow::Borrowed(arguments),
ToolPayload::Custom { input } => Cow::Borrowed(input),
ToolPayload::LocalShell { params } => Cow::Owned(params.command.join(" ")),
ToolPayload::Mcp { raw_arguments, .. } => Cow::Borrowed(raw_arguments),
}
}
}
pub trait ToolOutput: Send {
fn log_preview(&self) -> String;
fn success_for_logging(&self) -> bool;
fn into_response(self, call_id: &str, payload: &ToolPayload) -> ResponseInputItem;
}
pub struct McpToolOutput {
pub result: Result<CallToolResult, String>,
}
impl ToolOutput for McpToolOutput {
fn log_preview(&self) -> String {
format!("{:?}", self.result)
}
fn success_for_logging(&self) -> bool {
self.result.is_ok()
}
fn into_response(self, call_id: &str, _payload: &ToolPayload) -> ResponseInputItem {
let Self { result } = self;
ResponseInputItem::McpToolCallOutput {
call_id: call_id.to_string(),
result,
}
}
}
pub struct FunctionToolOutput {
pub body: Vec<FunctionCallOutputContentItem>,
pub success: Option<bool>,
}
impl FunctionToolOutput {
pub fn from_text(text: String, success: Option<bool>) -> Self {
Self {
body: vec![FunctionCallOutputContentItem::InputText { text }],
success,
}
}
pub fn from_content(
content: Vec<FunctionCallOutputContentItem>,
success: Option<bool>,
) -> Self {
Self {
body: content,
success,
}
}
}
impl ToolOutput for FunctionToolOutput {
fn log_preview(&self) -> String {
telemetry_preview(
&function_call_output_content_items_to_text(&self.body).unwrap_or_default(),
)
}
fn success_for_logging(&self) -> bool {
self.success.unwrap_or(true)
}
fn into_response(self, call_id: &str, payload: &ToolPayload) -> ResponseInputItem {
let Self { body, success } = self;
function_tool_response(call_id, payload, body, success)
}
}
fn function_tool_response(
call_id: &str,
payload: &ToolPayload,
body: Vec<FunctionCallOutputContentItem>,
success: Option<bool>,
) -> ResponseInputItem {
let body = match body.as_slice() {
[FunctionCallOutputContentItem::InputText { text }] => {
FunctionCallOutputBody::Text(text.clone())
}
_ => FunctionCallOutputBody::ContentItems(body),
};
if matches!(payload, ToolPayload::Custom { .. }) {
return ResponseInputItem::CustomToolCallOutput {
call_id: call_id.to_string(),
output: FunctionCallOutputPayload { body, success },
};
}
ResponseInputItem::FunctionCallOutput {
call_id: call_id.to_string(),
output: FunctionCallOutputPayload { body, success },
}
}
fn telemetry_preview(content: &str) -> String {
let truncated_slice = take_bytes_at_char_boundary(content, TELEMETRY_PREVIEW_MAX_BYTES);
let truncated_by_bytes = truncated_slice.len() < content.len();
let mut preview = String::new();
let mut lines_iter = truncated_slice.lines();
for idx in 0..TELEMETRY_PREVIEW_MAX_LINES {
match lines_iter.next() {
Some(line) => {
if idx > 0 {
preview.push('\n');
}
preview.push_str(line);
}
None => break,
}
}
let truncated_by_lines = lines_iter.next().is_some();
if !truncated_by_bytes && !truncated_by_lines {
return content.to_string();
}
if preview.len() < truncated_slice.len()
&& truncated_slice
.as_bytes()
.get(preview.len())
.is_some_and(|byte| *byte == b'\n')
{
preview.push('\n');
}
if !preview.is_empty() && !preview.ends_with('\n') {
preview.push('\n');
}
preview.push_str(TELEMETRY_PREVIEW_TRUNCATION_NOTICE);
preview
}
#[cfg(test)]
mod tests {
use super::*;
use pretty_assertions::assert_eq;
#[test]
fn custom_tool_calls_should_roundtrip_as_custom_outputs() {
let payload = ToolPayload::Custom {
input: "patch".to_string(),
};
let response = FunctionToolOutput::from_text("patched".to_string(), Some(true))
.into_response("call-42", &payload);
match response {
ResponseInputItem::CustomToolCallOutput { call_id, output } => {
assert_eq!(call_id, "call-42");
assert_eq!(output.content_items(), None);
assert_eq!(output.body.to_text().as_deref(), Some("patched"));
assert_eq!(output.success, Some(true));
}
other => panic!("expected CustomToolCallOutput, got {other:?}"),
}
}
#[test]
fn function_payloads_remain_function_outputs() {
let payload = ToolPayload::Function {
arguments: "{}".to_string(),
};
let response = FunctionToolOutput::from_text("ok".to_string(), Some(true))
.into_response("fn-1", &payload);
match response {
ResponseInputItem::FunctionCallOutput { call_id, output } => {
assert_eq!(call_id, "fn-1");
assert_eq!(output.content_items(), None);
assert_eq!(output.body.to_text().as_deref(), Some("ok"));
assert_eq!(output.success, Some(true));
}
other => panic!("expected FunctionCallOutput, got {other:?}"),
}
}
#[test]
fn custom_tool_calls_can_derive_text_from_content_items() {
let payload = ToolPayload::Custom {
input: "patch".to_string(),
};
let response = FunctionToolOutput::from_content(
vec![
FunctionCallOutputContentItem::InputText {
text: "line 1".to_string(),
},
FunctionCallOutputContentItem::InputImage {
image_url: "data:image/png;base64,AAA".to_string(),
detail: None,
},
FunctionCallOutputContentItem::InputText {
text: "line 2".to_string(),
},
],
Some(true),
)
.into_response("call-99", &payload);
match response {
ResponseInputItem::CustomToolCallOutput { call_id, output } => {
let expected = vec![
FunctionCallOutputContentItem::InputText {
text: "line 1".to_string(),
},
FunctionCallOutputContentItem::InputImage {
image_url: "data:image/png;base64,AAA".to_string(),
detail: None,
},
FunctionCallOutputContentItem::InputText {
text: "line 2".to_string(),
},
];
assert_eq!(call_id, "call-99");
assert_eq!(output.content_items(), Some(expected.as_slice()));
assert_eq!(output.body.to_text().as_deref(), Some("line 1\nline 2"));
assert_eq!(output.success, Some(true));
}
other => panic!("expected CustomToolCallOutput, got {other:?}"),
}
}
#[test]
fn log_preview_uses_content_items_when_plain_text_is_missing() {
let output = FunctionToolOutput::from_content(
vec![FunctionCallOutputContentItem::InputText {
text: "preview".to_string(),
}],
Some(true),
);
assert_eq!(output.log_preview(), "preview");
assert_eq!(
function_call_output_content_items_to_text(&output.body),
Some("preview".to_string())
);
}
#[test]
fn telemetry_preview_returns_original_within_limits() {
let content = "short output";
assert_eq!(telemetry_preview(content), content);
}
#[test]
fn telemetry_preview_truncates_by_bytes() {
let content = "x".repeat(TELEMETRY_PREVIEW_MAX_BYTES + 8);
let preview = telemetry_preview(&content);
assert!(preview.contains(TELEMETRY_PREVIEW_TRUNCATION_NOTICE));
assert!(
preview.len()
<= TELEMETRY_PREVIEW_MAX_BYTES + TELEMETRY_PREVIEW_TRUNCATION_NOTICE.len() + 1
);
}
#[test]
fn telemetry_preview_truncates_by_lines() {
let content = (0..(TELEMETRY_PREVIEW_MAX_LINES + 5))
.map(|idx| format!("line {idx}"))
.collect::<Vec<_>>()
.join("\n");
let preview = telemetry_preview(&content);
let lines: Vec<&str> = preview.lines().collect();
assert!(lines.len() <= TELEMETRY_PREVIEW_MAX_LINES + 1);
assert_eq!(lines.last(), Some(&TELEMETRY_PREVIEW_TRUNCATION_NOTICE));
}
}