objectiveai-api 2.0.5

ObjectiveAI API Server
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54

//! Wire types for the `objectiveai-codex-sdk-runner` stdio NDJSON
//! protocol. Mirrors `claude_agent_sdk::stdio` exactly — one long-lived
//! Python subprocess multiplexes N concurrent codex `Thread.run_streamed`
//! calls over a single (stdin, stdout, stderr) triple. The caller tags
//! every `run` request with a string `id`; every line emitted on stdout
//! and stderr carries the same `id` so the caller can demultiplex N
//! concurrent streams.
//!
//! ## Wire shapes
//!
//! Every line is one JSON object terminated by `\n`. By convention,
//! **`type` is the first field, `id` is the second field** on every
//! tagged line.
//!
//! Stdout (per-request, both variants of [`StdioOutput`] carry `id`):
//!
//! ```text
//! {"type":"event","id":"<id>","event":<ThreadEvent>}
//! {"type":"end","id":"<id>","status":"ok"}
//! {"type":"end","id":"<id>","status":"error","error":"<msg>"}
//! ```
//!
//! Stderr (per-request `diag` carries `id`; process-fatal does not):
//!
//! ```text
//! {"type":"diag","id":"<id>","level":"info|warn|error","message":"..."}
//! {"type":"fatal","message":"..."}
//! ```
//!
//! In-flight cancellation is intentionally absent: codex's `Thread`
//! doesn't expose a stop point that doesn't leave a billing event
//! unaccounted for, so once a `run` is sent the runner runs it to
//! natural completion.

mod run_params;
pub use run_params::*;
mod runner;
pub use runner::*;
mod runner_error;
pub use runner_error::*;
mod runner_stream;
pub use runner_stream::*;
mod runner_update;
pub use runner_update::*;
mod stdio_diag_level;
pub use stdio_diag_level::*;
mod stdio_end_status;
pub use stdio_end_status::*;
mod stdio_error;
pub use stdio_error::*;
mod stdio_input;
pub use stdio_input::*;
mod stdio_output;
pub use stdio_output::*;