config-reference.md Codex Docs, 2026-03-17 18:24 UTC → 2026-05-07 17:08 UTC

8 8 

9For sandbox and approval keys (`approval_policy`, `sandbox_mode`, and `sandbox_workspace_write.*`), pair this reference with [Sandbox and approvals](https://developers.openai.com/codex/agent-approvals-security#sandbox-and-approvals), [Protected paths in writable roots](https://developers.openai.com/codex/agent-approvals-security#protected-paths-in-writable-roots), and [Network access](https://developers.openai.com/codex/agent-approvals-security#network-access).9For sandbox and approval keys (`approval_policy`, `sandbox_mode`, and `sandbox_workspace_write.*`), pair this reference with [Sandbox and approvals](https://developers.openai.com/codex/agent-approvals-security#sandbox-and-approvals), [Protected paths in writable roots](https://developers.openai.com/codex/agent-approvals-security#protected-paths-in-writable-roots), and [Network access](https://developers.openai.com/codex/agent-approvals-security#network-access).

10 10 

11| Key | Type / Values | Details |11<ConfigTable

12| --- | --- | --- |12 options={[

13| `agents.<name>.config_file` | `string (path)` | Path to a TOML config layer for that role; relative paths resolve from the config file that declares the role. |13 {

14| `agents.<name>.description` | `string` | Role guidance shown to Codex when choosing and spawning that agent type. |14 key: "model",

15| `agents.<name>.nickname_candidates` | `array<string>` | Optional pool of display nicknames for spawned agents in that role. |15 type: "string",

16| `agents.job_max_runtime_seconds` | `number` | Default per-worker timeout for `spawn_agents_on_csv` jobs. When unset, the tool falls back to 1800 seconds per worker. |16 description: "Model to use (e.g., `gpt-5.5`).",

17| `agents.max_depth` | `number` | Maximum nesting depth allowed for spawned agent threads (root sessions start at depth 0; default: 1). |17 },

18| `agents.max_threads` | `number` | Maximum number of agent threads that can be open concurrently. Defaults to `6` when unset. |18 {

19| `allow_login_shell` | `boolean` | Allow shell-based tools to use login-shell semantics. Defaults to `true`; when `false`, `login = true` requests are rejected and omitted `login` defaults to non-login shells. |19 key: "review_model",

20| `analytics.enabled` | `boolean` | Enable or disable analytics for this machine/profile. When unset, the client default applies. |20 type: "string",

21| `approval_policy` | `untrusted | on-request | never | { reject = { sandbox_approval = bool, rules = bool, mcp_elicitations = bool } }` | Controls when Codex pauses for approval before executing commands. You can also use `approval_policy = { reject = { ... } }` to auto-reject specific prompt categories while keeping other prompts interactive. `on-failure` is deprecated; use `on-request` for interactive runs or `never` for non-interactive runs. |21 description:

22| `approval_policy.reject.mcp_elicitations` | `boolean` | When `true`, MCP elicitation prompts are auto-rejected instead of shown to the user. |22 "Optional model override used by `/review` (defaults to the current session model).",

23| `approval_policy.reject.rules` | `boolean` | When `true`, approvals triggered by execpolicy `prompt` rules are auto-rejected. |23 },

24| `approval_policy.reject.sandbox_approval` | `boolean` | When `true`, sandbox escalation approval prompts are auto-rejected. |24 {

25| `apps._default.destructive_enabled` | `boolean` | Default allow/deny for app tools with `destructive_hint = true`. |25 key: "model_provider",

26| `apps._default.enabled` | `boolean` | Default app enabled state for all apps unless overridden per app. |26 type: "string",

27| `apps._default.open_world_enabled` | `boolean` | Default allow/deny for app tools with `open_world_hint = true`. |27 description: "Provider id from `model_providers` (default: `openai`).",

28| `apps.<id>.default_tools_approval_mode` | `auto | prompt | approve` | Default approval behavior for tools in this app unless a per-tool override exists. |28 },

29| `apps.<id>.default_tools_enabled` | `boolean` | Default enabled state for tools in this app unless a per-tool override exists. |29 {

30| `apps.<id>.destructive_enabled` | `boolean` | Allow or block tools in this app that advertise `destructive_hint = true`. |30 key: "openai_base_url",

31| `apps.<id>.enabled` | `boolean` | Enable or disable a specific app/connector by id (default: true). |31 type: "string",

32| `apps.<id>.open_world_enabled` | `boolean` | Allow or block tools in this app that advertise `open_world_hint = true`. |32 description:

33| `apps.<id>.tools.<tool>.approval_mode` | `auto | prompt | approve` | Per-tool approval behavior override for a single app tool. |33 "Base URL override for the built-in `openai` model provider.",

34| `apps.<id>.tools.<tool>.enabled` | `boolean` | Per-tool enabled override for an app tool (for example `repos/list`). |34 },

35| `background_terminal_max_timeout` | `number` | Maximum poll window in milliseconds for empty `write_stdin` polls (background terminal polling). Default: `300000` (5 minutes). Replaces the older `background_terminal_timeout` key. |35 {

36| `chatgpt_base_url` | `string` | Override the base URL used during the ChatGPT login flow. |36 key: "model_context_window",

37| `check_for_update_on_startup` | `boolean` | Check for Codex updates on startup (set to false only when updates are centrally managed). |37 type: "number",

38| `cli_auth_credentials_store` | `file | keyring | auto` | Control where the CLI stores cached credentials (file-based auth.json vs OS keychain). |38 description: "Context window tokens available to the active model.",

39| `commit_attribution` | `string` | Override the commit co-author trailer text. Set an empty string to disable automatic attribution. |39 },

40| `compact_prompt` | `string` | Inline override for the history compaction prompt. |40 {

41| `developer_instructions` | `string` | Additional developer instructions injected into the session (optional). |41 key: "model_auto_compact_token_limit",

42| `disable_paste_burst` | `boolean` | Disable burst-paste detection in the TUI. |42 type: "number",

43| `experimental_compact_prompt_file` | `string (path)` | Load the compaction prompt override from a file (experimental). |43 description:

44| `experimental_use_unified_exec_tool` | `boolean` | Legacy name for enabling unified exec; prefer `[features].unified_exec` or `codex --enable unified_exec`. |44 "Token threshold that triggers automatic history compaction (unset uses model defaults).",

45| `features.apps` | `boolean` | Enable ChatGPT Apps/connectors support (experimental). |45 },

46| `features.apps_mcp_gateway` | `boolean` | Route Apps MCP calls through the OpenAI connectors MCP gateway (`https://api.openai.com/v1/connectors/mcp/`) instead of legacy routing (experimental). |46 {

47| `features.artifact` | `boolean` | Enable native artifact tools such as slides and spreadsheets (under development). |47 key: "model_catalog_json",

48| `features.child_agents_md` | `boolean` | Append AGENTS.md scope/precedence guidance even when no AGENTS.md is present (experimental). |48 type: "string (path)",

49| `features.collaboration_modes` | `boolean` | Legacy toggle for collaboration modes. Plan and default modes are available in current builds without setting this key. |49 description:

50| `features.default_mode_request_user_input` | `boolean` | Allow `request_user_input` in default collaboration mode (under development; off by default). |50 "Optional path to a JSON model catalog loaded on startup. Profile-level `profiles.<name>.model_catalog_json` can override this per profile.",

51| `features.elevated_windows_sandbox` | `boolean` | Legacy toggle for an earlier elevated Windows sandbox rollout. Current builds do not use it. |51 },

52| `features.enable_request_compression` | `boolean` | Compress streaming request bodies with zstd when supported (stable; on by default). |52 {

53| `features.experimental_windows_sandbox` | `boolean` | Legacy toggle for an earlier Windows sandbox rollout. Current builds do not use it. |53 key: "oss_provider",

54| `features.fast_mode` | `boolean` | Enable Fast mode selection and the `service_tier = "fast"` path (stable; on by default). |54 type: "lmstudio | ollama",

55| `features.image_detail_original` | `boolean` | Allow image outputs with `detail = "original"` on supported models (under development). |55 description:

56| `features.image_generation` | `boolean` | Enable the built-in image generation tool (under development). |56 "Default local provider used when running with `--oss` (defaults to prompting if unset).",

57| `features.personality` | `boolean` | Enable personality selection controls (stable; on by default). |57 },

58| `features.powershell_utf8` | `boolean` | Force PowerShell UTF-8 output. Enabled by default on Windows and off elsewhere. |58 {

59| `features.prevent_idle_sleep` | `boolean` | Prevent the machine from sleeping while a turn is actively running (experimental; off by default). |59 key: "approval_policy",

60| `features.remote_models` | `boolean` | Legacy toggle for an older remote-model readiness flow. Current builds do not use it. |60 type: "untrusted | on-request | never | { granular = { sandbox_approval = bool, rules = bool, mcp_elicitations = bool, request_permissions = bool, skill_approval = bool } }",

61| `features.request_rule` | `boolean` | Legacy toggle for Smart approvals. Current builds include this behavior by default, so most users can leave this unset. |61 description:

62| `features.responses_websockets` | `boolean` | Prefer the Responses API WebSocket transport for supported providers (under development). |62 "Controls when Codex pauses for approval before executing commands. You can also use `approval_policy = { granular = { ... } }` to allow or auto-reject specific prompt categories while keeping other prompts interactive. `on-failure` is deprecated; use `on-request` for interactive runs or `never` for non-interactive runs.",

63| `features.responses_websockets_v2` | `boolean` | Enable Responses API WebSocket v2 mode (under development). |63 },

64| `features.runtime_metrics` | `boolean` | Show runtime metrics summary in TUI turn separators (experimental). |64 {

65| `features.search_tool` | `boolean` | Legacy toggle for an older Apps discovery flow. Current builds do not use it. |65 key: "approval_policy.granular.sandbox_approval",

66| `features.shell_snapshot` | `boolean` | Snapshot shell environment to speed up repeated commands (stable; on by default). |66 type: "boolean",

67| `features.shell_tool` | `boolean` | Enable the default `shell` tool for running commands (stable; on by default). |67 description:

68| `features.skill_env_var_dependency_prompt` | `boolean` | Prompt for missing skill environment-variable dependencies (under development). |68 "When `true`, sandbox escalation approval prompts are allowed to surface.",

69| `features.skill_mcp_dependency_install` | `boolean` | Allow prompting and installing missing MCP dependencies for skills (stable; on by default). |69 },

70| `features.sqlite` | `boolean` | Enable SQLite-backed state persistence (stable; on by default). |70 {

71| `features.steer` | `boolean` | Legacy toggle from an earlier Enter/Tab steering rollout. Current builds always use the current steering behavior. |71 key: "approval_policy.granular.rules",

72| `features.undo` | `boolean` | Enable undo support (stable; off by default). |72 type: "boolean",

73| `features.unified_exec` | `boolean` | Use the unified PTY-backed exec tool (stable; enabled by default except on Windows). |73 description:

74| `features.use_linux_sandbox_bwrap` | `boolean` | Use the bubblewrap-based Linux sandbox pipeline (experimental; off by default). |74 "When `true`, approvals triggered by execpolicy `prompt` rules are allowed to surface.",

75| `features.web_search` | `boolean` | Deprecated legacy toggle; prefer the top-level `web_search` setting. |75 },

76| `features.web_search_cached` | `boolean` | Deprecated legacy toggle. When `web_search` is unset, true maps to `web_search = "cached"`. |76 {

77| `features.web_search_request` | `boolean` | Deprecated legacy toggle. When `web_search` is unset, true maps to `web_search = "live"`. |77 key: "approval_policy.granular.mcp_elicitations",

78| `feedback.enabled` | `boolean` | Enable feedback submission via `/feedback` across Codex surfaces (default: true). |78 type: "boolean",

79| `file_opener` | `vscode | vscode-insiders | windsurf | cursor | none` | URI scheme used to open citations from Codex output (default: `vscode`). |79 description:

80| `forced_chatgpt_workspace_id` | `string (uuid)` | Limit ChatGPT logins to a specific workspace identifier. |80 "When `true`, MCP elicitation prompts are allowed to surface instead of being auto-rejected.",

81| `forced_login_method` | `chatgpt | api` | Restrict Codex to a specific authentication method. |81 },

82| `hide_agent_reasoning` | `boolean` | Suppress reasoning events in both the TUI and `codex exec` output. |82 {

83| `history.max_bytes` | `number` | If set, caps the history file size in bytes by dropping oldest entries. |83 key: "approval_policy.granular.request_permissions",

84| `history.persistence` | `save-all | none` | Control whether Codex saves session transcripts to history.jsonl. |84 type: "boolean",

85| `instructions` | `string` | Reserved for future use; prefer `model_instructions_file` or `AGENTS.md`. |85 description:

86| `log_dir` | `string (path)` | Directory where Codex writes log files (for example `codex-tui.log`); defaults to `$CODEX_HOME/log`. |86 "When `true`, prompts from the `request_permissions` tool are allowed to surface.",

87| `mcp_oauth_callback_port` | `integer` | Optional fixed port for the local HTTP callback server used during MCP OAuth login. When unset, Codex binds to an ephemeral port chosen by the OS. |87 },

88| `mcp_oauth_callback_url` | `string` | Optional redirect URI override for MCP OAuth login (for example, a devbox ingress URL). `mcp_oauth_callback_port` still controls the callback listener port. |88 {

89| `mcp_oauth_credentials_store` | `auto | file | keyring` | Preferred store for MCP OAuth credentials. |89 key: "approval_policy.granular.skill_approval",

90| `mcp_servers.<id>.args` | `array<string>` | Arguments passed to the MCP stdio server command. |90 type: "boolean",

91| `mcp_servers.<id>.bearer_token_env_var` | `string` | Environment variable sourcing the bearer token for an MCP HTTP server. |91 description:

92| `mcp_servers.<id>.command` | `string` | Launcher command for an MCP stdio server. |92 "When `true`, skill-script approval prompts are allowed to surface.",

93| `mcp_servers.<id>.cwd` | `string` | Working directory for the MCP stdio server process. |93 },

94| `mcp_servers.<id>.disabled_tools` | `array<string>` | Deny list applied after `enabled_tools` for the MCP server. |94 {

95| `mcp_servers.<id>.enabled` | `boolean` | Disable an MCP server without removing its configuration. |95 key: "approvals_reviewer",

96| `mcp_servers.<id>.enabled_tools` | `array<string>` | Allow list of tool names exposed by the MCP server. |96 type: "user | auto_review",

97| `mcp_servers.<id>.env` | `map<string,string>` | Environment variables forwarded to the MCP stdio server. |97 description:

98| `mcp_servers.<id>.env_http_headers` | `map<string,string>` | HTTP headers populated from environment variables for an MCP HTTP server. |98 "Who reviews eligible approval prompts under `on-request` or granular approval policies. Defaults to `user`; `auto_review` uses the reviewer subagent. This setting doesn't change sandboxing or review actions already allowed inside the sandbox.",

99| `mcp_servers.<id>.env_vars` | `array<string>` | Additional environment variables to whitelist for an MCP stdio server. |99 },

100| `mcp_servers.<id>.http_headers` | `map<string,string>` | Static HTTP headers included with each MCP HTTP request. |100 {

101| `mcp_servers.<id>.oauth_resource` | `string` | Optional RFC 8707 OAuth resource parameter to include during MCP login. |101 key: "auto_review.policy",

102| `mcp_servers.<id>.required` | `boolean` | When true, fail startup/resume if this enabled MCP server cannot initialize. |102 type: "string",

103| `mcp_servers.<id>.scopes` | `array<string>` | OAuth scopes to request when authenticating to that MCP server. |103 description:

104| `mcp_servers.<id>.startup_timeout_ms` | `number` | Alias for `startup_timeout_sec` in milliseconds. |104 "Local Markdown policy instructions for automatic review. Managed `guardian_policy_config` takes precedence. Blank values are ignored.",

105| `mcp_servers.<id>.startup_timeout_sec` | `number` | Override the default 10s startup timeout for an MCP server. |105 },

106| `mcp_servers.<id>.tool_timeout_sec` | `number` | Override the default 60s per-tool timeout for an MCP server. |106 {

107| `mcp_servers.<id>.url` | `string` | Endpoint for an MCP streamable HTTP server. |107 key: "allow_login_shell",

108| `model` | `string` | Model to use (e.g., `gpt-5-codex`). |108 type: "boolean",

109| `model_auto_compact_token_limit` | `number` | Token threshold that triggers automatic history compaction (unset uses model defaults). |109 description:

110| `model_catalog_json` | `string (path)` | Optional path to a JSON model catalog loaded on startup. Profile-level `profiles.<name>.model_catalog_json` can override this per profile. |110 "Allow shell-based tools to use login-shell semantics. Defaults to `true`; when `false`, `login = true` requests are rejected and omitted `login` defaults to non-login shells.",

111| `model_context_window` | `number` | Context window tokens available to the active model. |111 },

112| `model_instructions_file` | `string (path)` | Replacement for built-in instructions instead of `AGENTS.md`. |112 {

113| `model_provider` | `string` | Provider id from `model_providers` (default: `openai`). |113 key: "sandbox_mode",

114| `model_providers.<id>.base_url` | `string` | API base URL for the model provider. |114 type: "read-only | workspace-write | danger-full-access",

115| `model_providers.<id>.env_http_headers` | `map<string,string>` | HTTP headers populated from environment variables when present. |115 description:

116| `model_providers.<id>.env_key` | `string` | Environment variable supplying the provider API key. |116 "Sandbox policy for filesystem and network access during command execution.",

117| `model_providers.<id>.env_key_instructions` | `string` | Optional setup guidance for the provider API key. |117 },

118| `model_providers.<id>.experimental_bearer_token` | `string` | Direct bearer token for the provider (discouraged; use `env_key`). |118 {

119| `model_providers.<id>.http_headers` | `map<string,string>` | Static HTTP headers added to provider requests. |119 key: "sandbox_workspace_write.writable_roots",

120| `model_providers.<id>.name` | `string` | Display name for a custom model provider. |120 type: "array<string>",

121| `model_providers.<id>.query_params` | `map<string,string>` | Extra query parameters appended to provider requests. |121 description:

122| `model_providers.<id>.request_max_retries` | `number` | Retry count for HTTP requests to the provider (default: 4). |122 'Additional writable roots when `sandbox_mode = "workspace-write"`.',

123| `model_providers.<id>.requires_openai_auth` | `boolean` | The provider uses OpenAI authentication (defaults to false). |123 },

124| `model_providers.<id>.stream_idle_timeout_ms` | `number` | Idle timeout for SSE streams in milliseconds (default: 300000). |124 {

125| `model_providers.<id>.stream_max_retries` | `number` | Retry count for SSE streaming interruptions (default: 5). |125 key: "sandbox_workspace_write.network_access",

126| `model_providers.<id>.supports_websockets` | `boolean` | Whether that provider supports the Responses API WebSocket transport. |126 type: "boolean",

127| `model_providers.<id>.wire_api` | `responses` | Protocol used by the provider. `responses` is the only supported value, and it is the default when omitted. |127 description:

128| `model_reasoning_effort` | `minimal | low | medium | high | xhigh` | Adjust reasoning effort for supported models (Responses API only; `xhigh` is model-dependent). |128 "Allow outbound network access inside the workspace-write sandbox.",

129| `model_reasoning_summary` | `auto | concise | detailed | none` | Select reasoning summary detail or disable summaries entirely. |129 },

130| `model_supports_reasoning_summaries` | `boolean` | Force Codex to send or not send reasoning metadata. |130 {

131| `model_verbosity` | `low | medium | high` | Optional GPT-5 Responses API verbosity override; when unset, the selected model/preset default is used. |131 key: "sandbox_workspace_write.exclude_tmpdir_env_var",

132| `notice.hide_full_access_warning` | `boolean` | Track acknowledgement of the full access warning prompt. |132 type: "boolean",

133| `notice.hide_gpt-5.1-codex-max_migration_prompt` | `boolean` | Track acknowledgement of the gpt-5.1-codex-max migration prompt. |133 description:

134| `notice.hide_gpt5_1_migration_prompt` | `boolean` | Track acknowledgement of the GPT-5.1 migration prompt. |134 "Exclude `$TMPDIR` from writable roots in workspace-write mode.",

135| `notice.hide_rate_limit_model_nudge` | `boolean` | Track opt-out of the rate limit model switch reminder. |135 },

136| `notice.hide_world_writable_warning` | `boolean` | Track acknowledgement of the Windows world-writable directories warning. |136 {

137| `notice.model_migrations` | `map<string,string>` | Track acknowledged model migrations as old->new mappings. |137 key: "sandbox_workspace_write.exclude_slash_tmp",

138| `notify` | `array<string>` | Command invoked for notifications; receives a JSON payload from Codex. |138 type: "boolean",

139| `oss_provider` | `lmstudio | ollama` | Default local provider used when running with `--oss` (defaults to prompting if unset). |139 description:

140| `otel.environment` | `string` | Environment tag applied to emitted OpenTelemetry events (default: `dev`). |140 "Exclude `/tmp` from writable roots in workspace-write mode.",

141| `otel.exporter` | `none | otlp-http | otlp-grpc` | Select the OpenTelemetry exporter and provide any endpoint metadata. |141 },

142| `otel.exporter.<id>.endpoint` | `string` | Exporter endpoint for OTEL logs. |142 {

143| `otel.exporter.<id>.headers` | `map<string,string>` | Static headers included with OTEL exporter requests. |143 key: "windows.sandbox",

144| `otel.exporter.<id>.protocol` | `binary | json` | Protocol used by the OTLP/HTTP exporter. |144 type: "unelevated | elevated",

145| `otel.exporter.<id>.tls.ca-certificate` | `string` | CA certificate path for OTEL exporter TLS. |145 description:

146| `otel.exporter.<id>.tls.client-certificate` | `string` | Client certificate path for OTEL exporter TLS. |146 "Windows-only native sandbox mode when running Codex natively on Windows.",

147| `otel.exporter.<id>.tls.client-private-key` | `string` | Client private key path for OTEL exporter TLS. |147 },

148| `otel.log_user_prompt` | `boolean` | Opt in to exporting raw user prompts with OpenTelemetry logs. |148 {

149| `otel.metrics_exporter` | `none | statsig | otlp-http | otlp-grpc` | Select the OpenTelemetry metrics exporter (defaults to `statsig`). |149 key: "windows.sandbox_private_desktop",

150| `otel.trace_exporter` | `none | otlp-http | otlp-grpc` | Select the OpenTelemetry trace exporter and provide any endpoint metadata. |150 type: "boolean",

151| `otel.trace_exporter.<id>.endpoint` | `string` | Trace exporter endpoint for OTEL logs. |151 description:

152| `otel.trace_exporter.<id>.headers` | `map<string,string>` | Static headers included with OTEL trace exporter requests. |152 "Run the final sandboxed child process on a private desktop by default on native Windows. Set `false` only for compatibility with the older `Winsta0\\\\Default` behavior.",

153| `otel.trace_exporter.<id>.protocol` | `binary | json` | Protocol used by the OTLP/HTTP trace exporter. |153 },

154| `otel.trace_exporter.<id>.tls.ca-certificate` | `string` | CA certificate path for OTEL trace exporter TLS. |154 {

155| `otel.trace_exporter.<id>.tls.client-certificate` | `string` | Client certificate path for OTEL trace exporter TLS. |155 key: "notify",

156| `otel.trace_exporter.<id>.tls.client-private-key` | `string` | Client private key path for OTEL trace exporter TLS. |156 type: "array<string>",

157| `permissions.network.admin_url` | `string` | Admin endpoint for the managed network proxy. |157 description:

158| `permissions.network.allow_local_binding` | `boolean` | Permit local bind/listen operations through the managed proxy. |158 "Command invoked for notifications; receives a JSON payload from Codex.",

159| `permissions.network.allow_unix_sockets` | `array<string>` | Allowlist of Unix socket paths permitted through the managed proxy. |159 },

160| `permissions.network.allow_upstream_proxy` | `boolean` | Allow the managed proxy to chain to another upstream proxy. |160 {

161| `permissions.network.allowed_domains` | `array<string>` | Allowlist of domains permitted through the managed proxy. |161 key: "check_for_update_on_startup",

162| `permissions.network.dangerously_allow_all_unix_sockets` | `boolean` | Allow the proxy to use arbitrary Unix sockets instead of the default restricted set. |162 type: "boolean",

163| `permissions.network.dangerously_allow_non_loopback_admin` | `boolean` | Permit non-loopback bind addresses for the managed proxy admin listener. |163 description:

164| `permissions.network.dangerously_allow_non_loopback_proxy` | `boolean` | Permit non-loopback bind addresses for the managed proxy listener. |164 "Check for Codex updates on startup (set to false only when updates are centrally managed).",

165| `permissions.network.denied_domains` | `array<string>` | Denylist of domains blocked by the managed proxy. |165 },

166| `permissions.network.enable_socks5` | `boolean` | Expose a SOCKS5 listener from the managed network proxy. |166 {

167| `permissions.network.enable_socks5_udp` | `boolean` | Allow UDP over the SOCKS5 listener when enabled. |167 key: "feedback.enabled",

168| `permissions.network.enabled` | `boolean` | Enable the managed network proxy configuration for subprocesses. |168 type: "boolean",

169| `permissions.network.mode` | `limited | full` | Network proxy mode used for subprocess traffic. |169 description:

170| `permissions.network.proxy_url` | `string` | HTTP proxy endpoint used by the managed network proxy. |170 "Enable feedback submission via `/feedback` across Codex surfaces (default: true).",

171| `permissions.network.socks_url` | `string` | SOCKS5 proxy endpoint used by the managed network proxy. |171 },

172| `personality` | `none | friendly | pragmatic` | Default communication style for models that advertise `supportsPersonality`; can be overridden per thread/turn or via `/personality`. |172 {

173| `plan_mode_reasoning_effort` | `none | minimal | low | medium | high | xhigh` | Plan-mode-specific reasoning override. When unset, Plan mode uses its built-in preset default. |173 key: "analytics.enabled",

174| `profile` | `string` | Default profile applied at startup (equivalent to `--profile`). |174 type: "boolean",

175| `profiles.<name>.*` | `various` | Profile-scoped overrides for any of the supported configuration keys. |175 description:

176| `profiles.<name>.analytics.enabled` | `boolean` | Profile-scoped analytics enablement override. |176 "Enable or disable analytics for this machine/profile. When unset, the client default applies.",

177| `profiles.<name>.experimental_use_unified_exec_tool` | `boolean` | Legacy name for enabling unified exec; prefer `[features].unified_exec`. |177 },

178| `profiles.<name>.model_catalog_json` | `string (path)` | Profile-scoped model catalog JSON path override (applied on startup only; overrides the top-level `model_catalog_json` for that profile). |178 {

179| `profiles.<name>.model_instructions_file` | `string (path)` | Profile-scoped replacement for the built-in instruction file. |179 key: "instructions",

180| `profiles.<name>.oss_provider` | `lmstudio | ollama` | Profile-scoped OSS provider for `--oss` sessions. |180 type: "string",

181| `profiles.<name>.personality` | `none | friendly | pragmatic` | Profile-scoped communication style override for supported models. |181 description:

182| `profiles.<name>.plan_mode_reasoning_effort` | `none | minimal | low | medium | high | xhigh` | Profile-scoped Plan-mode reasoning override. |182 "Reserved for future use; prefer `model_instructions_file` or `AGENTS.md`.",

183| `profiles.<name>.service_tier` | `flex | fast` | Profile-scoped service tier preference for new turns. |183 },

184| `profiles.<name>.tools_view_image` | `boolean` | Enable or disable the `view_image` tool in that profile. |184 {

185| `profiles.<name>.web_search` | `disabled | cached | live` | Profile-scoped web search mode override (default: `"cached"`). |185 key: "developer_instructions",

186| `profiles.<name>.windows.sandbox` | `unelevated | elevated` | Profile-scoped Windows sandbox mode override. |186 type: "string",

187| `project_doc_fallback_filenames` | `array<string>` | Additional filenames to try when `AGENTS.md` is missing. |187 description:

188| `project_doc_max_bytes` | `number` | Maximum bytes read from `AGENTS.md` when building project instructions. |188 "Additional developer instructions injected into the session (optional).",

189| `project_root_markers` | `array<string>` | List of project root marker filenames; used when searching parent directories for the project root. |189 },

190| `projects.<path>.trust_level` | `string` | Mark a project or worktree as trusted or untrusted (`"trusted"` | `"untrusted"`). Untrusted projects skip project-scoped `.codex/` layers. |190 {

191| `review_model` | `string` | Optional model override used by `/review` (defaults to the current session model). |191 key: "log_dir",

192| `sandbox_mode` | `read-only | workspace-write | danger-full-access` | Sandbox policy for filesystem and network access during command execution. |192 type: "string (path)",

193| `sandbox_workspace_write.exclude_slash_tmp` | `boolean` | Exclude `/tmp` from writable roots in workspace-write mode. |193 description:

194| `sandbox_workspace_write.exclude_tmpdir_env_var` | `boolean` | Exclude `$TMPDIR` from writable roots in workspace-write mode. |194 "Directory where Codex writes log files (for example `codex-tui.log`); defaults to `$CODEX_HOME/log`.",

195| `sandbox_workspace_write.network_access` | `boolean` | Allow outbound network access inside the workspace-write sandbox. |195 },

196| `sandbox_workspace_write.writable_roots` | `array<string>` | Additional writable roots when `sandbox_mode = "workspace-write"`. |196 {

197| `service_tier` | `flex | fast` | Preferred service tier for new turns. `fast` is honored only when the `features.fast_mode` gate is enabled. |197 key: "sqlite_home",

198| `shell_environment_policy.exclude` | `array<string>` | Glob patterns for removing environment variables after the defaults. |198 type: "string (path)",

199| `shell_environment_policy.experimental_use_profile` | `boolean` | Use the user shell profile when spawning subprocesses. |199 description:

200| `shell_environment_policy.ignore_default_excludes` | `boolean` | Keep variables containing KEY/SECRET/TOKEN before other filters run. |200 "Directory where Codex stores the SQLite-backed state DB used by agent jobs and other resumable runtime state.",

201| `shell_environment_policy.include_only` | `array<string>` | Whitelist of patterns; when set only matching variables are kept. |201 },

202| `shell_environment_policy.inherit` | `all | core | none` | Baseline environment inheritance when spawning subprocesses. |202 {

203| `shell_environment_policy.set` | `map<string,string>` | Explicit environment overrides injected into every subprocess. |203 key: "compact_prompt",

204| `show_raw_agent_reasoning` | `boolean` | Surface raw reasoning content when the active model emits it. |204 type: "string",

205| `skills.config` | `array<object>` | Per-skill enablement overrides stored in config.toml. |205 description: "Inline override for the history compaction prompt.",

206| `skills.config.<index>.enabled` | `boolean` | Enable or disable the referenced skill. |206 },

207| `skills.config.<index>.path` | `string (path)` | Path to a skill folder containing `SKILL.md`. |207 {

208| `sqlite_home` | `string (path)` | Directory where Codex stores the SQLite-backed state DB used by agent jobs and other resumable runtime state. |208 key: "commit_attribution",

209| `suppress_unstable_features_warning` | `boolean` | Suppress the warning that appears when under-development feature flags are enabled. |209 type: "string",

210| `tool_output_token_limit` | `number` | Token budget for storing individual tool/function outputs in history. |210 description:

211| `tools.view_image` | `boolean` | Enable the local-image attachment tool `view_image`. |211 "Override the commit co-author trailer text. Set an empty string to disable automatic attribution.",

212| `tools.web_search` | `boolean` | Deprecated legacy toggle for web search; prefer the top-level `web_search` setting. |212 },

213| `tui` | `table` | TUI-specific options such as enabling inline desktop notifications. |213 {

214| `tui.alternate_screen` | `auto | always | never` | Control alternate screen usage for the TUI (default: auto; auto skips it in Zellij to preserve scrollback). |214 key: "model_instructions_file",

215| `tui.animations` | `boolean` | Enable terminal animations (welcome screen, shimmer, spinner) (default: true). |215 type: "string (path)",

216| `tui.model_availability_nux.<model>` | `integer` | Internal startup-tooltip state keyed by model slug. |216 description:

217| `tui.notification_method` | `auto | osc9 | bel` | Notification method for unfocused terminal notifications (default: auto). |217 "Replacement for built-in instructions instead of `AGENTS.md`.",

218| `tui.notifications` | `boolean | array<string>` | Enable TUI notifications; optionally restrict to specific event types. |218 },

219| `tui.show_tooltips` | `boolean` | Show onboarding tooltips in the TUI welcome screen (default: true). |219 {

220| `tui.status_line` | `array<string> | null` | Ordered list of TUI footer status-line item identifiers. `null` disables the status line. |220 key: "personality",

221| `tui.theme` | `string` | Syntax-highlighting theme override (kebab-case theme name). |221 type: "none | friendly | pragmatic",

222| `web_search` | `disabled | cached | live` | Web search mode (default: `"cached"`; cached uses an OpenAI-maintained index and does not fetch live pages; if you use `--yolo` or another full access sandbox setting, it defaults to `"live"`). Use `"live"` to fetch the most recent data from the web, or `"disabled"` to remove the tool. |222 description:

223| `windows_wsl_setup_acknowledged` | `boolean` | Track Windows onboarding acknowledgement (Windows only). |223 "Default communication style for models that advertise `supportsPersonality`; can be overridden per thread/turn or via `/personality`.",

224| `windows.sandbox` | `unelevated | elevated` | Windows-only native sandbox mode when running Codex natively on Windows. |224 },

225 225 {

226Key226 key: "service_tier",

227 227 type: "flex | fast",

228`agents.<name>.config_file`228 description: "Preferred service tier for new turns.",

229 229 },

230Type / Values230 {

231 231 key: "experimental_compact_prompt_file",

232`string (path)`232 type: "string (path)",

233 233 description:

234Details234 "Load the compaction prompt override from a file (experimental).",

235 235 },

236Path to a TOML config layer for that role; relative paths resolve from the config file that declares the role.236 {

237 237 key: "skills.config",

238Key238 type: "array<object>",

239 239 description: "Per-skill enablement overrides stored in config.toml.",

240`agents.<name>.description`240 },

241 241 {

242Type / Values242 key: "skills.config.<index>.path",

243 243 type: "string (path)",

244`string`244 description: "Path to a skill folder containing `SKILL.md`.",

245 245 },

246Details246 {

247 247 key: "skills.config.<index>.enabled",

248Role guidance shown to Codex when choosing and spawning that agent type.248 type: "boolean",

249 249 description: "Enable or disable the referenced skill.",

250Key250 },

251 251 {

252`agents.<name>.nickname_candidates`252 key: "apps.<id>.enabled",

253 253 type: "boolean",

254Type / Values254 description:

255 255 "Enable or disable a specific app/connector by id (default: true).",

256`array<string>`256 },

257 257 {

258Details258 key: "apps._default.enabled",

259 259 type: "boolean",

260Optional pool of display nicknames for spawned agents in that role.260 description:

261 261 "Default app enabled state for all apps unless overridden per app.",

262Key262 },

263 263 {

264`agents.job_max_runtime_seconds`264 key: "apps._default.destructive_enabled",

265 265 type: "boolean",

266Type / Values266 description:

267 267 "Default allow/deny for app tools with `destructive_hint = true`.",

268`number`268 },

269 269 {

270Details270 key: "apps._default.open_world_enabled",

271 271 type: "boolean",

272Default per-worker timeout for `spawn_agents_on_csv` jobs. When unset, the tool falls back to 1800 seconds per worker.272 description:

273 273 "Default allow/deny for app tools with `open_world_hint = true`.",

274Key274 },

275 275 {

276`agents.max_depth`276 key: "apps.<id>.destructive_enabled",

277 277 type: "boolean",

278Type / Values278 description:

279 279 "Allow or block tools in this app that advertise `destructive_hint = true`.",

280`number`280 },

281 281 {

282Details282 key: "apps.<id>.open_world_enabled",

283 283 type: "boolean",

284Maximum nesting depth allowed for spawned agent threads (root sessions start at depth 0; default: 1).284 description:

285 285 "Allow or block tools in this app that advertise `open_world_hint = true`.",

286Key286 },

287 287 {

288`agents.max_threads`288 key: "apps.<id>.default_tools_enabled",

289 289 type: "boolean",

290Type / Values290 description:

291 291 "Default enabled state for tools in this app unless a per-tool override exists.",

292`number`292 },

293 293 {

294Details294 key: "apps.<id>.default_tools_approval_mode",

295 295 type: "auto | prompt | approve",

296Maximum number of agent threads that can be open concurrently. Defaults to `6` when unset.296 description:

297 297 "Default approval behavior for tools in this app unless a per-tool override exists.",

298Key298 },

299 299 {

300`allow_login_shell`300 key: "apps.<id>.tools.<tool>.enabled",

301 301 type: "boolean",

302Type / Values302 description:

303 303 "Per-tool enabled override for an app tool (for example `repos/list`).",

304`boolean`304 },

305 305 {

306Details306 key: "apps.<id>.tools.<tool>.approval_mode",

307 307 type: "auto | prompt | approve",

308Allow shell-based tools to use login-shell semantics. Defaults to `true`; when `false`, `login = true` requests are rejected and omitted `login` defaults to non-login shells.308 description: "Per-tool approval behavior override for a single app tool.",

309 309 },

310Key310 {

311 311 key: "tool_suggest.discoverables",

312`analytics.enabled`312 type: "array<table>",

313 313 description:

314Type / Values314 'Allow tool suggestions for additional discoverable connectors or plugins. Each entry uses `type = "connector"` or `"plugin"` and an `id`.',

315 315 },

316`boolean`316 {

317 317 key: "tool_suggest.disabled_tools",

318Details318 type: "array<table>",

319 319 description:

320Enable or disable analytics for this machine/profile. When unset, the client default applies.320 'Disable suggestions for specific discoverable connectors or plugins. Each entry uses `type = "connector"` or `"plugin"` and an `id`.',

321 321 },

322Key322 {

323 323 key: "features.apps",

324`approval_policy`324 type: "boolean",

325 325 description: "Enable ChatGPT Apps/connectors support (experimental).",

326Type / Values326 },

327 327 {

328`untrusted | on-request | never | { reject = { sandbox_approval = bool, rules = bool, mcp_elicitations = bool } }`328 key: "features.codex_hooks",

329 329 type: "boolean",

330Details330 description:

331 331 "Enable lifecycle hooks loaded from `hooks.json` or inline `[hooks]` config.",

332Controls when Codex pauses for approval before executing commands. You can also use `approval_policy = { reject = { ... } }` to auto-reject specific prompt categories while keeping other prompts interactive. `on-failure` is deprecated; use `on-request` for interactive runs or `never` for non-interactive runs.332 },

333 333 {

334Key334 key: "hooks",

335 335 type: "table",

336`approval_policy.reject.mcp_elicitations`336 description:

337 337 "Lifecycle hooks configured inline in `config.toml`. Uses the same event schema as `hooks.json`; see the Hooks guide for examples and supported events.",

338Type / Values338 },

339 339 {

340`boolean`340 key: "features.memories",

341 341 type: "boolean",

342Details342 description: "Enable [Memories](https://developers.openai.com/codex/memories) (off by default).",

343 343 },

344When `true`, MCP elicitation prompts are auto-rejected instead of shown to the user.344 {

345 345 key: "mcp_servers.<id>.command",

346Key346 type: "string",

347 347 description: "Launcher command for an MCP stdio server.",

348`approval_policy.reject.rules`348 },

349 349 {

350Type / Values350 key: "mcp_servers.<id>.args",

351 351 type: "array<string>",

352`boolean`352 description: "Arguments passed to the MCP stdio server command.",

353 353 },

354Details354 {

355 355 key: "mcp_servers.<id>.env",

356When `true`, approvals triggered by execpolicy `prompt` rules are auto-rejected.356 type: "map<string,string>",

357 357 description: "Environment variables forwarded to the MCP stdio server.",

358Key358 },

359 359 {

360`approval_policy.reject.sandbox_approval`360 key: "mcp_servers.<id>.env_vars",

361 361 type: 'array<string | { name = string, source = "local" | "remote" }>',

362Type / Values362 description:

363 363 'Additional environment variables to whitelist for an MCP stdio server. String entries default to `source = "local"`; use `source = "remote"` only with executor-backed remote stdio.',

364`boolean`364 },

365 365 {

366Details366 key: "mcp_servers.<id>.cwd",

367 367 type: "string",

368When `true`, sandbox escalation approval prompts are auto-rejected.368 description: "Working directory for the MCP stdio server process.",

369 369 },

370Key370 {

371 371 key: "mcp_servers.<id>.url",

372`apps._default.destructive_enabled`372 type: "string",

373 373 description: "Endpoint for an MCP streamable HTTP server.",

374Type / Values374 },

375 375 {

376`boolean`376 key: "mcp_servers.<id>.bearer_token_env_var",

377 377 type: "string",

378Details378 description:

379 379 "Environment variable sourcing the bearer token for an MCP HTTP server.",

380Default allow/deny for app tools with `destructive_hint = true`.380 },

381 381 {

382Key382 key: "mcp_servers.<id>.http_headers",

383 383 type: "map<string,string>",

384`apps._default.enabled`384 description: "Static HTTP headers included with each MCP HTTP request.",

385 385 },

386Type / Values386 {

387 387 key: "mcp_servers.<id>.env_http_headers",

388`boolean`388 type: "map<string,string>",

389 389 description:

390Details390 "HTTP headers populated from environment variables for an MCP HTTP server.",

391 391 },

392Default app enabled state for all apps unless overridden per app.392 {

393 393 key: "mcp_servers.<id>.enabled",

394Key394 type: "boolean",

395 395 description: "Disable an MCP server without removing its configuration.",

396`apps._default.open_world_enabled`396 },

397 397 {

398Type / Values398 key: "mcp_servers.<id>.required",

399 399 type: "boolean",

400`boolean`400 description:

401 401 "When true, fail startup/resume if this enabled MCP server cannot initialize.",

402Details402 },

403 403 {

404Default allow/deny for app tools with `open_world_hint = true`.404 key: "mcp_servers.<id>.startup_timeout_sec",

405 405 type: "number",

406Key406 description:

407 407 "Override the default 10s startup timeout for an MCP server.",

408`apps.<id>.default_tools_approval_mode`408 },

409 409 {

410Type / Values410 key: "mcp_servers.<id>.startup_timeout_ms",

411 411 type: "number",

412`auto | prompt | approve`412 description: "Alias for `startup_timeout_sec` in milliseconds.",

413 413 },

414Details414 {

415 415 key: "mcp_servers.<id>.tool_timeout_sec",

416Default approval behavior for tools in this app unless a per-tool override exists.416 type: "number",

417 417 description:

418Key418 "Override the default 60s per-tool timeout for an MCP server.",

419 419 },

420`apps.<id>.default_tools_enabled`420 {

421 421 key: "mcp_servers.<id>.enabled_tools",

422Type / Values422 type: "array<string>",

423 423 description: "Allow list of tool names exposed by the MCP server.",

424`boolean`424 },

425 425 {

426Details426 key: "mcp_servers.<id>.disabled_tools",

427 427 type: "array<string>",

428Default enabled state for tools in this app unless a per-tool override exists.428 description:

429 429 "Deny list applied after `enabled_tools` for the MCP server.",

430Key430 },

431 431 {

432`apps.<id>.destructive_enabled`432 key: "mcp_servers.<id>.scopes",

433 433 type: "array<string>",

434Type / Values434 description:

435 435 "OAuth scopes to request when authenticating to that MCP server.",

436`boolean`436 },

437 437 {

438Details438 key: "mcp_servers.<id>.oauth_resource",

439 439 type: "string",

440Allow or block tools in this app that advertise `destructive_hint = true`.440 description:

441 441 "Optional RFC 8707 OAuth resource parameter to include during MCP login.",

442Key442 },

443 443 {

444`apps.<id>.enabled`444 key: "mcp_servers.<id>.experimental_environment",

445 445 type: "local | remote",

446Type / Values446 description:

447 447 "Experimental placement for an MCP server. `remote` starts stdio servers through a remote executor environment; streamable HTTP remote placement is not implemented.",

448`boolean`448 },

449 449 {

450Details450 key: "agents.max_threads",

451 451 type: "number",

452Enable or disable a specific app/connector by id (default: true).452 description:

453 453 "Maximum number of agent threads that can be open concurrently. Defaults to `6` when unset.",

454Key454 },

455 455 {

456`apps.<id>.open_world_enabled`456 key: "agents.max_depth",

457 457 type: "number",

458Type / Values458 description:

459 459 "Maximum nesting depth allowed for spawned agent threads (root sessions start at depth 0; default: 1).",

460`boolean`460 },

461 461 {

462Details462 key: "agents.job_max_runtime_seconds",

463 463 type: "number",

464Allow or block tools in this app that advertise `open_world_hint = true`.464 description:

465 465 "Default per-worker timeout for `spawn_agents_on_csv` jobs. When unset, the tool falls back to 1800 seconds per worker.",

466Key466 },

467 467 {

468`apps.<id>.tools.<tool>.approval_mode`468 key: "agents.<name>.description",

469 469 type: "string",

470Type / Values470 description:

471 471 "Role guidance shown to Codex when choosing and spawning that agent type.",

472`auto | prompt | approve`472 },

473 473 {

474Details474 key: "agents.<name>.config_file",

475 475 type: "string (path)",

476Per-tool approval behavior override for a single app tool.476 description:

477 477 "Path to a TOML config layer for that role; relative paths resolve from the config file that declares the role.",

478Key478 },

479 479 {

480`apps.<id>.tools.<tool>.enabled`480 key: "agents.<name>.nickname_candidates",

481 481 type: "array<string>",

482Type / Values482 description:

483 483 "Optional pool of display nicknames for spawned agents in that role.",

484`boolean`484 },

485 485 {

486Details486 key: "memories.generate_memories",

487 487 type: "boolean",

488Per-tool enabled override for an app tool (for example `repos/list`).488 description:

489 489 "When `false`, newly created threads are not stored as memory-generation inputs. Defaults to `true`.",

490Key490 },

491 491 {

492`background_terminal_max_timeout`492 key: "memories.use_memories",

493 493 type: "boolean",

494Type / Values494 description:

495 495 "When `false`, Codex skips injecting existing memories into future sessions. Defaults to `true`.",

496`number`496 },

497 497 {

498Details498 key: "memories.disable_on_external_context",

499 499 type: "boolean",

500Maximum poll window in milliseconds for empty `write_stdin` polls (background terminal polling). Default: `300000` (5 minutes). Replaces the older `background_terminal_timeout` key.500 description:

501 501 "When `true`, threads that use external context such as MCP tool calls, web search, or tool search are kept out of memory generation. Defaults to `false`. Legacy alias: `memories.no_memories_if_mcp_or_web_search`.",

502Key502 },

503 503 {

504`chatgpt_base_url`504 key: "memories.max_raw_memories_for_consolidation",

505 505 type: "number",

506Type / Values506 description:

507 507 "Maximum recent raw memories retained for global consolidation. Defaults to `256` and is capped at `4096`.",

508`string`508 },

509 509 {

510Details510 key: "memories.max_unused_days",

511 511 type: "number",

512Override the base URL used during the ChatGPT login flow.512 description:

513 513 "Maximum days since a memory was last used before it becomes ineligible for consolidation. Defaults to `30` and is clamped to `0`-`365`.",

514Key514 },

515 515 {

516`check_for_update_on_startup`516 key: "memories.max_rollout_age_days",

517 517 type: "number",

518Type / Values518 description:

519 519 "Maximum age of threads considered for memory generation. Defaults to `30` and is clamped to `0`-`90`.",

520`boolean`520 },

521 521 {

522Details522 key: "memories.max_rollouts_per_startup",

523 523 type: "number",

524Check for Codex updates on startup (set to false only when updates are centrally managed).524 description:

525 525 "Maximum rollout candidates processed per startup pass. Defaults to `16` and is capped at `128`.",

526Key526 },

527 527 {

528`cli_auth_credentials_store`528 key: "memories.min_rollout_idle_hours",

529 529 type: "number",

530Type / Values530 description:

531 531 "Minimum idle time before a thread is considered for memory generation. Defaults to `6` and is clamped to `1`-`48`.",

532`file | keyring | auto`532 },

533 533 {

534Details534 key: "memories.min_rate_limit_remaining_percent",

535 535 type: "number",

536Control where the CLI stores cached credentials (file-based auth.json vs OS keychain).536 description:

537 537 "Minimum remaining percentage required in Codex rate-limit windows before memory generation starts. Defaults to `25` and is clamped to `0`-`100`.",

538Key538 },

539 539 {

540`commit_attribution`540 key: "memories.extract_model",

541 541 type: "string",

542Type / Values542 description: "Optional model override for per-thread memory extraction.",

543 543 },

544`string`544 {

545 545 key: "memories.consolidation_model",

546Details546 type: "string",

547 547 description: "Optional model override for global memory consolidation.",

548Override the commit co-author trailer text. Set an empty string to disable automatic attribution.548 },

549 549 {

550Key550 key: "features.unified_exec",

551 551 type: "boolean",

552`compact_prompt`552 description:

553 553 "Use the unified PTY-backed exec tool (stable; enabled by default except on Windows).",

554Type / Values554 },

555 555 {

556`string`556 key: "features.shell_snapshot",

557 557 type: "boolean",

558Details558 description:

559 559 "Snapshot shell environment to speed up repeated commands (stable; on by default).",

560Inline override for the history compaction prompt.560 },

561 561 {

562Key562 key: "features.undo",

563 563 type: "boolean",

564`developer_instructions`564 description: "Enable undo support (stable; off by default).",

565 565 },

566Type / Values566 {

567 567 key: "features.multi_agent",

568`string`568 type: "boolean",

569 569 description:

570Details570 "Enable multi-agent collaboration tools (`spawn_agent`, `send_input`, `resume_agent`, `wait_agent`, and `close_agent`) (stable; on by default).",

571 571 },

572Additional developer instructions injected into the session (optional).572 {

573 573 key: "features.personality",

574Key574 type: "boolean",

575 575 description:

576`disable_paste_burst`576 "Enable personality selection controls (stable; on by default).",

577 577 },

578Type / Values578 {

579 579 key: "features.web_search",

580`boolean`580 type: "boolean",

581 581 description:

582Details582 "Deprecated legacy toggle; prefer the top-level `web_search` setting.",

583 583 },

584Disable burst-paste detection in the TUI.584 {

585 585 key: "features.web_search_cached",

586Key586 type: "boolean",

587 587 description:

588`experimental_compact_prompt_file`588 'Deprecated legacy toggle. When `web_search` is unset, true maps to `web_search = "cached"`.',

589 589 },

590Type / Values590 {

591 591 key: "features.web_search_request",

592`string (path)`592 type: "boolean",

593 593 description:

594Details594 'Deprecated legacy toggle. When `web_search` is unset, true maps to `web_search = "live"`.',

595 595 },

596Load the compaction prompt override from a file (experimental).596 {

597 597 key: "features.shell_tool",

598Key598 type: "boolean",

599 599 description:

600`experimental_use_unified_exec_tool`600 "Enable the default `shell` tool for running commands (stable; on by default).",

601 601 },

602Type / Values602 {

603 603 key: "features.enable_request_compression",

604`boolean`604 type: "boolean",

605 605 description:

606Details606 "Compress streaming request bodies with zstd when supported (stable; on by default).",

607 607 },

608Legacy name for enabling unified exec; prefer `[features].unified_exec` or `codex --enable unified_exec`.608 {

609 609 key: "features.skill_mcp_dependency_install",

610Key610 type: "boolean",

611 611 description:

612`features.apps`612 "Allow prompting and installing missing MCP dependencies for skills (stable; on by default).",

613 613 },

614Type / Values614 {

615 615 key: "features.fast_mode",

616`boolean`616 type: "boolean",

617 617 description:

618Details618 'Enable Fast mode selection and the `service_tier = "fast"` path (stable; on by default).',

619 619 },

620Enable ChatGPT Apps/connectors support (experimental).620 {

621 621 key: "features.prevent_idle_sleep",

622Key622 type: "boolean",

623 623 description:

624`features.apps_mcp_gateway`624 "Prevent the machine from sleeping while a turn is actively running (experimental; off by default).",

625 625 },

626Type / Values626 {

627 627 key: "suppress_unstable_features_warning",

628`boolean`628 type: "boolean",

629 629 description:

630Details630 "Suppress the warning that appears when under-development feature flags are enabled.",

631 631 },

632Route Apps MCP calls through the OpenAI connectors MCP gateway (`https://api.openai.com/v1/connectors/mcp/`) instead of legacy routing (experimental).632 {

633 633 key: "model_providers.<id>",

634Key634 type: "table",

635 635 description:

636`features.artifact`636 "Custom provider definition. Built-in provider IDs (`openai`, `ollama`, and `lmstudio`) are reserved and cannot be overridden.",

637 637 },

638Type / Values638 {

639 639 key: "model_providers.<id>.name",

640`boolean`640 type: "string",

641 641 description: "Display name for a custom model provider.",

642Details642 },

643 643 {

644Enable native artifact tools such as slides and spreadsheets (under development).644 key: "model_providers.<id>.base_url",

645 645 type: "string",

646Key646 description: "API base URL for the model provider.",

647 647 },

648`features.child_agents_md`648 {

649 649 key: "model_providers.<id>.env_key",

650Type / Values650 type: "string",

651 651 description: "Environment variable supplying the provider API key.",

652`boolean`652 },

653 653 {

654Details654 key: "model_providers.<id>.env_key_instructions",

655 655 type: "string",

656Append AGENTS.md scope/precedence guidance even when no AGENTS.md is present (experimental).656 description: "Optional setup guidance for the provider API key.",

657 657 },

658Key658 {

659 659 key: "model_providers.<id>.experimental_bearer_token",

660`features.collaboration_modes`660 type: "string",

661 661 description:

662Type / Values662 "Direct bearer token for the provider (discouraged; use `env_key`).",

663 663 },

664`boolean`664 {

665 665 key: "model_providers.<id>.requires_openai_auth",

666Details666 type: "boolean",

667 667 description:

668Legacy toggle for collaboration modes. Plan and default modes are available in current builds without setting this key.668 "The provider uses OpenAI authentication (defaults to false).",

669 669 },

670Key670 {

671 671 key: "model_providers.<id>.wire_api",

672`features.default_mode_request_user_input`672 type: "responses",

673 673 description:

674Type / Values674 "Protocol used by the provider. `responses` is the only supported value, and it is the default when omitted.",

675 675 },

676`boolean`676 {

677 677 key: "model_providers.<id>.query_params",

678Details678 type: "map<string,string>",

679 679 description: "Extra query parameters appended to provider requests.",

680Allow `request_user_input` in default collaboration mode (under development; off by default).680 },

681 681 {

682Key682 key: "model_providers.<id>.http_headers",

683 683 type: "map<string,string>",

684`features.elevated_windows_sandbox`684 description: "Static HTTP headers added to provider requests.",

685 685 },

686Type / Values686 {

687 687 key: "model_providers.<id>.env_http_headers",

688`boolean`688 type: "map<string,string>",

689 689 description:

690Details690 "HTTP headers populated from environment variables when present.",

691 691 },

692Legacy toggle for an earlier elevated Windows sandbox rollout. Current builds do not use it.692 {

693 693 key: "model_providers.<id>.request_max_retries",

694Key694 type: "number",

695 695 description:

696`features.enable_request_compression`696 "Retry count for HTTP requests to the provider (default: 4).",

697 697 },

698Type / Values698 {

699 699 key: "model_providers.<id>.stream_max_retries",

700`boolean`700 type: "number",

701 701 description: "Retry count for SSE streaming interruptions (default: 5).",

702Details702 },

703 703 {

704Compress streaming request bodies with zstd when supported (stable; on by default).704 key: "model_providers.<id>.stream_idle_timeout_ms",

705 705 type: "number",

706Key706 description:

707 707 "Idle timeout for SSE streams in milliseconds (default: 300000).",

708`features.experimental_windows_sandbox`708 },

709 709 {

710Type / Values710 key: "model_providers.<id>.supports_websockets",

711 711 type: "boolean",

712`boolean`712 description:

713 713 "Whether that provider supports the Responses API WebSocket transport.",

714Details714 },

715 715 {

716Legacy toggle for an earlier Windows sandbox rollout. Current builds do not use it.716 key: "model_providers.<id>.auth",

717 717 type: "table",

718Key718 description:

719 719 "Command-backed bearer token configuration for a custom provider. Do not combine with `env_key`, `experimental_bearer_token`, or `requires_openai_auth`.",

720`features.fast_mode`720 },

721 721 {

722Type / Values722 key: "model_providers.<id>.auth.command",

723 723 type: "string",

724`boolean`724 description:

725 725 "Command to run when Codex needs a bearer token. The command must print the token to stdout.",

726Details726 },

727 727 {

728Enable Fast mode selection and the `service_tier = "fast"` path (stable; on by default).728 key: "model_providers.<id>.auth.args",

729 729 type: "array<string>",

730Key730 description: "Arguments passed to the token command.",

731 731 },

732`features.image_detail_original`732 {

733 733 key: "model_providers.<id>.auth.timeout_ms",

734Type / Values734 type: "number",

735 735 description:

736`boolean`736 "Maximum token command runtime in milliseconds (default: 5000).",

737 737 },

738Details738 {

739 739 key: "model_providers.<id>.auth.refresh_interval_ms",

740Allow image outputs with `detail = "original"` on supported models (under development).740 type: "number",

741 741 description:

742Key742 "How often Codex proactively refreshes the token in milliseconds (default: 300000). Set to `0` to refresh only after an authentication retry.",

743 743 },

744`features.image_generation`744 {

745 745 key: "model_providers.<id>.auth.cwd",

746Type / Values746 type: "string (path)",

747 747 description: "Working directory for the token command.",

748`boolean`748 },

749 749 {

750Details750 key: "model_providers.amazon-bedrock.aws.profile",

751 751 type: "string",

752Enable the built-in image generation tool (under development).752 description:

753 753 "AWS profile name used by the built-in `amazon-bedrock` provider.",

754Key754 },

755 755 {

756`features.personality`756 key: "model_providers.amazon-bedrock.aws.region",

757 757 type: "string",

758Type / Values758 description: "AWS region used by the built-in `amazon-bedrock` provider.",

759 759 },

760`boolean`760 {

761 761 key: "model_reasoning_effort",

762Details762 type: "minimal | low | medium | high | xhigh",

763 763 description:

764Enable personality selection controls (stable; on by default).764 "Adjust reasoning effort for supported models (Responses API only; `xhigh` is model-dependent).",

765 765 },

766Key766 {

767 767 key: "plan_mode_reasoning_effort",

768`features.powershell_utf8`768 type: "none | minimal | low | medium | high | xhigh",

769 769 description:

770Type / Values770 "Plan-mode-specific reasoning override. When unset, Plan mode uses its built-in preset default.",

771 771 },

772`boolean`772 {

773 773 key: "model_reasoning_summary",

774Details774 type: "auto | concise | detailed | none",

775 775 description:

776Force PowerShell UTF-8 output. Enabled by default on Windows and off elsewhere.776 "Select reasoning summary detail or disable summaries entirely.",

777 777 },

778Key778 {

779 779 key: "model_verbosity",

780`features.prevent_idle_sleep`780 type: "low | medium | high",

781 781 description:

782Type / Values782 "Optional GPT-5 Responses API verbosity override; when unset, the selected model/preset default is used.",

783 783 },

784`boolean`784 {

785 785 key: "model_supports_reasoning_summaries",

786Details786 type: "boolean",

787 787 description: "Force Codex to send or not send reasoning metadata.",

788Prevent the machine from sleeping while a turn is actively running (experimental; off by default).788 },

789 789 {

790Key790 key: "shell_environment_policy.inherit",

791 791 type: "all | core | none",

792`features.remote_models`792 description:

793 793 "Baseline environment inheritance when spawning subprocesses.",

794Type / Values794 },

795 795 {

796`boolean`796 key: "shell_environment_policy.ignore_default_excludes",

797 797 type: "boolean",

798Details798 description:

799 799 "Keep variables containing KEY/SECRET/TOKEN before other filters run.",

800Legacy toggle for an older remote-model readiness flow. Current builds do not use it.800 },

801 801 {

802Key802 key: "shell_environment_policy.exclude",

803 803 type: "array<string>",

804`features.request_rule`804 description:

805 805 "Glob patterns for removing environment variables after the defaults.",

806Type / Values806 },

807 807 {

808`boolean`808 key: "shell_environment_policy.include_only",

809 809 type: "array<string>",

810Details810 description:

811 811 "Whitelist of patterns; when set only matching variables are kept.",

812Legacy toggle for Smart approvals. Current builds include this behavior by default, so most users can leave this unset.812 },

813 813 {