Structs§
- Active
Permission Profile - Metadata for the named or implicit built-in permissions profile that
produced the active
PermissionProfile. - Additional
Permission Profile - Partial permission overlay used for per-command requests and approved session/turn grants.
- Base
Instructions - Base instructions for the model in a thread. Corresponds to the
instructionsfield in the ResponsesAPI. - File
System Permissions - Function
Call Output Payload - The payload we send back to OpenAI when reporting a tool call result.
- Internal
Chat Message Metadata Passthrough - Internal Responses API passthrough metadata copied into underlying chat messages.
- Legacy
Read Write Roots - Local
Shell Exec Action - Network
Permissions - Search
Tool Call Params - Shell
Command Tool Call Params - If the
nameof aResponseItem::FunctionCallisshell_command, theargumentsfield should deserialize to this struct.
Enums§
- Agent
Message Input Content - Content
Item - Function
Call Output Body - Function
Call Output Content Item - Responses API compatible content items that can be returned by a tool call. This is a subset of ContentItem with the types we support as function call outputs.
- Image
Detail - Local
Image Preparation - Local
Shell Action - Local
Shell Status - Managed
File System Permissions - Filesystem permissions for profiles where Codex owns sandbox construction.
- Message
Phase - Classifies an assistant message as interim commentary or final answer text.
- Permission
Profile - Canonical active runtime permissions for a conversation, turn, or command.
- Reasoning
Item Content - Reasoning
Item Reasoning Summary - Response
Input Item - Response
Item - Sandbox
Enforcement - Sandbox
Permissions - Controls the per-command sandbox override requested by a shell-like tool call.
- WebSearch
Action
Constants§
- BASE_
INSTRUCTIONS_ DEFAULT - BUILT_
IN_ PERMISSION_ PROFILE_ DANGER_ FULL_ ACCESS - Reserved identifier for the built-in full-access permission profile.
- BUILT_
IN_ PERMISSION_ PROFILE_ READ_ ONLY - Reserved identifier for the built-in read-only permission profile.
- BUILT_
IN_ PERMISSION_ PROFILE_ WORKSPACE - Reserved identifier for the built-in workspace-write permission profile.
- DEFAULT_
IMAGE_ DETAIL - VIEW_
IMAGE_ TOOL_ NAME
Functions§
- audio_
close_ tag_ text - audio_
open_ tag_ text - format_
allow_ prefixes - function_
call_ output_ content_ items_ to_ text - Converts structured function-call output content into plain text for human-readable surfaces.
- image_
close_ tag_ text - image_
open_ tag_ text - is_
audio_ close_ tag_ text - is_
audio_ open_ tag_ text - is_
image_ close_ tag_ text - is_
image_ open_ tag_ text - is_
local_ audio_ close_ tag_ text - is_
local_ audio_ open_ tag_ text - is_
local_ image_ close_ tag_ text - is_
local_ image_ open_ tag_ text - local_
audio_ label_ text - local_
audio_ open_ tag_ text_ with_ path - local_
image_ content_ items_ with_ label_ number - local_
image_ label_ text - local_
image_ open_ tag_ text_ with_ path - plaintext_
agent_ message_ content - Returns the locally readable text when an agent message is entirely plaintext.