config-reference.md Codex Docs, 2026-02-27 18:15 UTC → 2026-04-21 12:30 UTC

6 6 

7User-level configuration lives in `~/.codex/config.toml`. You can also add project-scoped overrides in `.codex/config.toml` files. Codex loads project-scoped config files only when you trust the project.7User-level configuration lives in `~/.codex/config.toml`. You can also add project-scoped overrides in `.codex/config.toml` files. Codex loads project-scoped config files only when you trust the project.

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/security#sandbox-and-approvals), [Protected paths in writable roots](https://developers.openai.com/codex/security#protected-paths-in-writable-roots), and [Network access](https://developers.openai.com/codex/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| Key | Type / Values | Details |

12| --- | --- | --- |12| --- | --- | --- |

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| `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. |

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

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

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. |

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

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

17| `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| `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. |

18| `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. |20| `analytics.enabled` | `boolean` | Enable or disable analytics for this machine/profile. When unset, the client default applies. |

19| `approval_policy.reject.mcp_elicitations` | `boolean` | When `true`, MCP elicitation prompts are auto-rejected instead of shown to the user. |21| `approval_policy` | `untrusted | on-request | never | { granular = { sandbox_approval = bool, rules = bool, mcp_elicitations = bool, request_permissions = bool, skill_approval = bool } }` | 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. |

20| `approval_policy.reject.rules` | `boolean` | When `true`, approvals triggered by execpolicy `prompt` rules are auto-rejected. |22| `approval_policy.granular.mcp_elicitations` | `boolean` | When `true`, MCP elicitation prompts are allowed to surface instead of being auto-rejected. |

21| `approval_policy.reject.sandbox_approval` | `boolean` | When `true`, sandbox escalation approval prompts are auto-rejected. |23| `approval_policy.granular.request_permissions` | `boolean` | When `true`, prompts from the `request_permissions` tool are allowed to surface. |

24| `approval_policy.granular.rules` | `boolean` | When `true`, approvals triggered by execpolicy `prompt` rules are allowed to surface. |

25| `approval_policy.granular.sandbox_approval` | `boolean` | When `true`, sandbox escalation approval prompts are allowed to surface. |

26| `approval_policy.granular.skill_approval` | `boolean` | When `true`, skill-script approval prompts are allowed to surface. |

27| `approvals_reviewer` | `user | guardian_subagent` | Select who reviews eligible approval prompts. Defaults to `user`; `guardian_subagent` routes supported reviews through the Guardian reviewer subagent. |

22| `apps._default.destructive_enabled` | `boolean` | Default allow/deny for app tools with `destructive_hint = true`. |28| `apps._default.destructive_enabled` | `boolean` | Default allow/deny for app tools with `destructive_hint = true`. |

23| `apps._default.enabled` | `boolean` | Default app enabled state for all apps unless overridden per app. |29| `apps._default.enabled` | `boolean` | Default app enabled state for all apps unless overridden per app. |

24| `apps._default.open_world_enabled` | `boolean` | Default allow/deny for app tools with `open_world_hint = true`. |30| `apps._default.open_world_enabled` | `boolean` | Default allow/deny for app tools with `open_world_hint = true`. |

33| `chatgpt_base_url` | `string` | Override the base URL used during the ChatGPT login flow. |39| `chatgpt_base_url` | `string` | Override the base URL used during the ChatGPT login flow. |

34| `check_for_update_on_startup` | `boolean` | Check for Codex updates on startup (set to false only when updates are centrally managed). |40| `check_for_update_on_startup` | `boolean` | Check for Codex updates on startup (set to false only when updates are centrally managed). |

35| `cli_auth_credentials_store` | `file | keyring | auto` | Control where the CLI stores cached credentials (file-based auth.json vs OS keychain). |41| `cli_auth_credentials_store` | `file | keyring | auto` | Control where the CLI stores cached credentials (file-based auth.json vs OS keychain). |

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

36| `compact_prompt` | `string` | Inline override for the history compaction prompt. |43| `compact_prompt` | `string` | Inline override for the history compaction prompt. |

44| `default_permissions` | `string` | Name of the default permissions profile to apply to sandboxed tool calls. |

37| `developer_instructions` | `string` | Additional developer instructions injected into the session (optional). |45| `developer_instructions` | `string` | Additional developer instructions injected into the session (optional). |

38| `disable_paste_burst` | `boolean` | Disable burst-paste detection in the TUI. |46| `disable_paste_burst` | `boolean` | Disable burst-paste detection in the TUI. |

39| `experimental_compact_prompt_file` | `string (path)` | Load the compaction prompt override from a file (experimental). |47| `experimental_compact_prompt_file` | `string (path)` | Load the compaction prompt override from a file (experimental). |

40| `experimental_use_freeform_apply_patch` | `boolean` | Legacy name for enabling freeform apply\_patch; prefer `[features].apply_patch_freeform` or `codex --enable apply_patch_freeform`. |

41| `experimental_use_unified_exec_tool` | `boolean` | Legacy name for enabling unified exec; prefer `[features].unified_exec` or `codex --enable unified_exec`. |48| `experimental_use_unified_exec_tool` | `boolean` | Legacy name for enabling unified exec; prefer `[features].unified_exec` or `codex --enable unified_exec`. |

42| `features.apply_patch_freeform` | `boolean` | Expose the freeform `apply_patch` tool (experimental). |

43| `features.apps` | `boolean` | Enable ChatGPT Apps/connectors support (experimental). |49| `features.apps` | `boolean` | Enable ChatGPT Apps/connectors support (experimental). |

44| `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). |50| `features.codex_hooks` | `boolean` | Enable lifecycle hooks loaded from `hooks.json` (under development; off by default). |

45| `features.child_agents_md` | `boolean` | Append AGENTS.md scope/precedence guidance even when no AGENTS.md is present (experimental). |51| `features.enable_request_compression` | `boolean` | Compress streaming request bodies with zstd when supported (stable; on by default). |

46| `features.collaboration_modes` | `boolean` | Enable collaboration modes such as plan mode (stable; on by default). |52| `features.fast_mode` | `boolean` | Enable Fast mode selection and the `service_tier = "fast"` path (stable; on by default). |

47| `features.multi_agent` | `boolean` | Enable multi-agent collaboration tools (`spawn\_agent`, `send\_input`, `resume\_agent`, `wait`, and `close\_agent`) (experimental; off by default). |53| `features.guardian_approval` | `boolean` | Route eligible approval requests through the guardian reviewer subagent (experimental; off by default). Use with `approvals_reviewer = "guardian_subagent"`. |

54| `features.memories` | `boolean` | Enable [Memories](https://developers.openai.com/codex/memories) (off by default). |

55| `features.multi_agent` | `boolean` | Enable multi-agent collaboration tools (`spawn_agent`, `send_input`, `resume_agent`, `wait_agent`, and `close_agent`) (stable; on by default). |

48| `features.personality` | `boolean` | Enable personality selection controls (stable; on by default). |56| `features.personality` | `boolean` | Enable personality selection controls (stable; on by default). |

49| `features.powershell_utf8` | `boolean` | Force PowerShell UTF-8 output (defaults to true). |57| `features.prevent_idle_sleep` | `boolean` | Prevent the machine from sleeping while a turn is actively running (experimental; off by default). |

50| `features.remote_models` | `boolean` | Refresh remote model list before showing readiness (experimental). |58| `features.shell_snapshot` | `boolean` | Snapshot shell environment to speed up repeated commands (stable; on by default). |

51| `features.request_rule` | `boolean` | Enable Smart approvals (`prefix_rule` suggestions on escalation requests; stable; on by default). |

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

53| `features.search_tool` | `boolean` | Enable `search_tool_bm25` for Apps tool discovery before invoking app MCP tools (experimental). |

54| `features.shell_snapshot` | `boolean` | Snapshot shell environment to speed up repeated commands (beta). |

55| `features.shell_tool` | `boolean` | Enable the default `shell` tool for running commands (stable; on by default). |59| `features.shell_tool` | `boolean` | Enable the default `shell` tool for running commands (stable; on by default). |

56| `features.unified_exec` | `boolean` | Use the unified PTY-backed exec tool (beta). |60| `features.skill_mcp_dependency_install` | `boolean` | Allow prompting and installing missing MCP dependencies for skills (stable; on by default). |

57| `features.use_linux_sandbox_bwrap` | `boolean` | Use the bubblewrap-based Linux sandbox pipeline (experimental; off by default). |61| `features.undo` | `boolean` | Enable undo support (stable; off by default). |

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

58| `features.web_search` | `boolean` | Deprecated legacy toggle; prefer the top-level `web_search` setting. |63| `features.web_search` | `boolean` | Deprecated legacy toggle; prefer the top-level `web_search` setting. |

59| `features.web_search_cached` | `boolean` | Deprecated legacy toggle. When `web_search` is unset, true maps to `web_search = "cached"`. |64| `features.web_search_cached` | `boolean` | Deprecated legacy toggle. When `web_search` is unset, true maps to `web_search = "cached"`. |

60| `features.web_search_request` | `boolean` | Deprecated legacy toggle. When `web_search` is unset, true maps to `web_search = "live"`. |65| `features.web_search_request` | `boolean` | Deprecated legacy toggle. When `web_search` is unset, true maps to `web_search = "live"`. |

65| `hide_agent_reasoning` | `boolean` | Suppress reasoning events in both the TUI and `codex exec` output. |70| `hide_agent_reasoning` | `boolean` | Suppress reasoning events in both the TUI and `codex exec` output. |

66| `history.max_bytes` | `number` | If set, caps the history file size in bytes by dropping oldest entries. |71| `history.max_bytes` | `number` | If set, caps the history file size in bytes by dropping oldest entries. |

67| `history.persistence` | `save-all | none` | Control whether Codex saves session transcripts to history.jsonl. |72| `history.persistence` | `save-all | none` | Control whether Codex saves session transcripts to history.jsonl. |

68| `include_apply_patch_tool` | `boolean` | Legacy name for enabling freeform apply\_patch; prefer `[features].apply_patch_freeform`. |

69| `instructions` | `string` | Reserved for future use; prefer `model_instructions_file` or `AGENTS.md`. |73| `instructions` | `string` | Reserved for future use; prefer `model_instructions_file` or `AGENTS.md`. |

70| `log_dir` | `string (path)` | Directory where Codex writes log files (for example `codex-tui.log`); defaults to `$CODEX_HOME/log`. |74| `log_dir` | `string (path)` | Directory where Codex writes log files (for example `codex-tui.log`); defaults to `$CODEX_HOME/log`. |

71| `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. |75| `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. |

80| `mcp_servers.<id>.enabled_tools` | `array<string>` | Allow list of tool names exposed by the MCP server. |84| `mcp_servers.<id>.enabled_tools` | `array<string>` | Allow list of tool names exposed by the MCP server. |

81| `mcp_servers.<id>.env` | `map<string,string>` | Environment variables forwarded to the MCP stdio server. |85| `mcp_servers.<id>.env` | `map<string,string>` | Environment variables forwarded to the MCP stdio server. |

82| `mcp_servers.<id>.env_http_headers` | `map<string,string>` | HTTP headers populated from environment variables for an MCP HTTP server. |86| `mcp_servers.<id>.env_http_headers` | `map<string,string>` | HTTP headers populated from environment variables for an MCP HTTP server. |

83| `mcp_servers.<id>.env_vars` | `array<string>` | Additional environment variables to whitelist for an MCP stdio server. |87| `mcp_servers.<id>.env_vars` | `array<string | { name = string, source = "local" | "remote" }>` | 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. |

88| `mcp_servers.<id>.experimental_environment` | `local | remote` | Experimental placement for an MCP server. `remote` starts stdio servers through a remote executor environment; streamable HTTP remote placement is not implemented. |

84| `mcp_servers.<id>.http_headers` | `map<string,string>` | Static HTTP headers included with each MCP HTTP request. |89| `mcp_servers.<id>.http_headers` | `map<string,string>` | Static HTTP headers included with each MCP HTTP request. |

90| `mcp_servers.<id>.oauth_resource` | `string` | Optional RFC 8707 OAuth resource parameter to include during MCP login. |

85| `mcp_servers.<id>.required` | `boolean` | When true, fail startup/resume if this enabled MCP server cannot initialize. |91| `mcp_servers.<id>.required` | `boolean` | When true, fail startup/resume if this enabled MCP server cannot initialize. |

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

86| `mcp_servers.<id>.startup_timeout_ms` | `number` | Alias for `startup_timeout_sec` in milliseconds. |93| `mcp_servers.<id>.startup_timeout_ms` | `number` | Alias for `startup_timeout_sec` in milliseconds. |

87| `mcp_servers.<id>.startup_timeout_sec` | `number` | Override the default 10s startup timeout for an MCP server. |94| `mcp_servers.<id>.startup_timeout_sec` | `number` | Override the default 10s startup timeout for an MCP server. |

88| `mcp_servers.<id>.tool_timeout_sec` | `number` | Override the default 60s per-tool timeout for an MCP server. |95| `mcp_servers.<id>.tool_timeout_sec` | `number` | Override the default 60s per-tool timeout for an MCP server. |

89| `mcp_servers.<id>.url` | `string` | Endpoint for an MCP streamable HTTP server. |96| `mcp_servers.<id>.url` | `string` | Endpoint for an MCP streamable HTTP server. |

90| `model` | `string` | Model to use (e.g., `gpt-5-codex`). |97| `memories.consolidation_model` | `string` | Optional model override for global memory consolidation. |

98| `memories.disable_on_external_context` | `boolean` | 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`. |

99| `memories.extract_model` | `string` | Optional model override for per-thread memory extraction. |

100| `memories.generate_memories` | `boolean` | When `false`, newly created threads are not stored as memory-generation inputs. Defaults to `true`. |

101| `memories.max_raw_memories_for_consolidation` | `number` | Maximum recent raw memories retained for global consolidation. Defaults to `256` and is capped at `4096`. |

102| `memories.max_rollout_age_days` | `number` | Maximum age of threads considered for memory generation. Defaults to `30` and is clamped to `0`-`90`. |

103| `memories.max_rollouts_per_startup` | `number` | Maximum rollout candidates processed per startup pass. Defaults to `16` and is capped at `128`. |

104| `memories.max_unused_days` | `number` | Maximum days since a memory was last used before it becomes ineligible for consolidation. Defaults to `30` and is clamped to `0`-`365`. |

105| `memories.min_rollout_idle_hours` | `number` | Minimum idle time before a thread is considered for memory generation. Defaults to `6` and is clamped to `1`-`48`. |

106| `memories.use_memories` | `boolean` | When `false`, Codex skips injecting existing memories into future sessions. Defaults to `true`. |

107| `model` | `string` | Model to use (e.g., `gpt-5.4`). |

91| `model_auto_compact_token_limit` | `number` | Token threshold that triggers automatic history compaction (unset uses model defaults). |108| `model_auto_compact_token_limit` | `number` | Token threshold that triggers automatic history compaction (unset uses model defaults). |

92| `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. |109| `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. |

93| `model_context_window` | `number` | Context window tokens available to the active model. |110| `model_context_window` | `number` | Context window tokens available to the active model. |

94| `model_instructions_file` | `string (path)` | Replacement for built-in instructions instead of `AGENTS.md`. |111| `model_instructions_file` | `string (path)` | Replacement for built-in instructions instead of `AGENTS.md`. |

95| `model_provider` | `string` | Provider id from `model_providers` (default: `openai`). |112| `model_provider` | `string` | Provider id from `model_providers` (default: `openai`). |

113| `model_providers.<id>` | `table` | Custom provider definition. Built-in provider IDs (`openai`, `ollama`, and `lmstudio`) are reserved and cannot be overridden. |

114| `model_providers.<id>.auth` | `table` | Command-backed bearer token configuration for a custom provider. Do not combine with `env_key`, `experimental_bearer_token`, or `requires_openai_auth`. |

115| `model_providers.<id>.auth.args` | `array<string>` | Arguments passed to the token command. |

116| `model_providers.<id>.auth.command` | `string` | Command to run when Codex needs a bearer token. The command must print the token to stdout. |

117| `model_providers.<id>.auth.cwd` | `string (path)` | Working directory for the token command. |

118| `model_providers.<id>.auth.refresh_interval_ms` | `number` | How often Codex proactively refreshes the token in milliseconds (default: 300000). Set to `0` to refresh only after an authentication retry. |

119| `model_providers.<id>.auth.timeout_ms` | `number` | Maximum token command runtime in milliseconds (default: 5000). |

96| `model_providers.<id>.base_url` | `string` | API base URL for the model provider. |120| `model_providers.<id>.base_url` | `string` | API base URL for the model provider. |

97| `model_providers.<id>.env_http_headers` | `map<string,string>` | HTTP headers populated from environment variables when present. |121| `model_providers.<id>.env_http_headers` | `map<string,string>` | HTTP headers populated from environment variables when present. |

98| `model_providers.<id>.env_key` | `string` | Environment variable supplying the provider API key. |122| `model_providers.<id>.env_key` | `string` | Environment variable supplying the provider API key. |

105| `model_providers.<id>.requires_openai_auth` | `boolean` | The provider uses OpenAI authentication (defaults to false). |129| `model_providers.<id>.requires_openai_auth` | `boolean` | The provider uses OpenAI authentication (defaults to false). |

106| `model_providers.<id>.stream_idle_timeout_ms` | `number` | Idle timeout for SSE streams in milliseconds (default: 300000). |130| `model_providers.<id>.stream_idle_timeout_ms` | `number` | Idle timeout for SSE streams in milliseconds (default: 300000). |

107| `model_providers.<id>.stream_max_retries` | `number` | Retry count for SSE streaming interruptions (default: 5). |131| `model_providers.<id>.stream_max_retries` | `number` | Retry count for SSE streaming interruptions (default: 5). |

108| `model_providers.<id>.wire_api` | `chat | responses` | Protocol used by the provider (defaults to `chat` if omitted). |132| `model_providers.<id>.supports_websockets` | `boolean` | Whether that provider supports the Responses API WebSocket transport. |

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

109| `model_reasoning_effort` | `minimal | low | medium | high | xhigh` | Adjust reasoning effort for supported models (Responses API only; `xhigh` is model-dependent). |134| `model_reasoning_effort` | `minimal | low | medium | high | xhigh` | Adjust reasoning effort for supported models (Responses API only; `xhigh` is model-dependent). |

110| `model_reasoning_summary` | `auto | concise | detailed | none` | Select reasoning summary detail or disable summaries entirely. |135| `model_reasoning_summary` | `auto | concise | detailed | none` | Select reasoning summary detail or disable summaries entirely. |

111| `model_supports_reasoning_summaries` | `boolean` | Force Codex to send or not send reasoning metadata. |136| `model_supports_reasoning_summaries` | `boolean` | Force Codex to send or not send reasoning metadata. |

112| `model_verbosity` | `low | medium | high` | Control GPT-5 Responses API verbosity (defaults to `medium`). |137| `model_verbosity` | `low | medium | high` | Optional GPT-5 Responses API verbosity override; when unset, the selected model/preset default is used. |

113| `notice.hide_full_access_warning` | `boolean` | Track acknowledgement of the full access warning prompt. |138| `notice.hide_full_access_warning` | `boolean` | Track acknowledgement of the full access warning prompt. |

114| `notice.hide_gpt-5.1-codex-max_migration_prompt` | `boolean` | Track acknowledgement of the gpt-5.1-codex-max migration prompt. |139| `notice.hide_gpt-5.1-codex-max_migration_prompt` | `boolean` | Track acknowledgement of the gpt-5.1-codex-max migration prompt. |

115| `notice.hide_gpt5_1_migration_prompt` | `boolean` | Track acknowledgement of the GPT-5.1 migration prompt. |140| `notice.hide_gpt5_1_migration_prompt` | `boolean` | Track acknowledgement of the GPT-5.1 migration prompt. |

117| `notice.hide_world_writable_warning` | `boolean` | Track acknowledgement of the Windows world-writable directories warning. |142| `notice.hide_world_writable_warning` | `boolean` | Track acknowledgement of the Windows world-writable directories warning. |

118| `notice.model_migrations` | `map<string,string>` | Track acknowledged model migrations as old->new mappings. |143| `notice.model_migrations` | `map<string,string>` | Track acknowledged model migrations as old->new mappings. |

119| `notify` | `array<string>` | Command invoked for notifications; receives a JSON payload from Codex. |144| `notify` | `array<string>` | Command invoked for notifications; receives a JSON payload from Codex. |

145| `openai_base_url` | `string` | Base URL override for the built-in `openai` model provider. |

120| `oss_provider` | `lmstudio | ollama` | Default local provider used when running with `--oss` (defaults to prompting if unset). |146| `oss_provider` | `lmstudio | ollama` | Default local provider used when running with `--oss` (defaults to prompting if unset). |

121| `otel.environment` | `string` | Environment tag applied to emitted OpenTelemetry events (default: `dev`). |147| `otel.environment` | `string` | Environment tag applied to emitted OpenTelemetry events (default: `dev`). |

122| `otel.exporter` | `none | otlp-http | otlp-grpc` | Select the OpenTelemetry exporter and provide any endpoint metadata. |148| `otel.exporter` | `none | otlp-http | otlp-grpc` | Select the OpenTelemetry exporter and provide any endpoint metadata. |

127| `otel.exporter.<id>.tls.client-certificate` | `string` | Client certificate path for OTEL exporter TLS. |153| `otel.exporter.<id>.tls.client-certificate` | `string` | Client certificate path for OTEL exporter TLS. |

128| `otel.exporter.<id>.tls.client-private-key` | `string` | Client private key path for OTEL exporter TLS. |154| `otel.exporter.<id>.tls.client-private-key` | `string` | Client private key path for OTEL exporter TLS. |

129| `otel.log_user_prompt` | `boolean` | Opt in to exporting raw user prompts with OpenTelemetry logs. |155| `otel.log_user_prompt` | `boolean` | Opt in to exporting raw user prompts with OpenTelemetry logs. |

156| `otel.metrics_exporter` | `none | statsig | otlp-http | otlp-grpc` | Select the OpenTelemetry metrics exporter (defaults to `statsig`). |

130| `otel.trace_exporter` | `none | otlp-http | otlp-grpc` | Select the OpenTelemetry trace exporter and provide any endpoint metadata. |157| `otel.trace_exporter` | `none | otlp-http | otlp-grpc` | Select the OpenTelemetry trace exporter and provide any endpoint metadata. |

131| `otel.trace_exporter.<id>.endpoint` | `string` | Trace exporter endpoint for OTEL logs. |158| `otel.trace_exporter.<id>.endpoint` | `string` | Trace exporter endpoint for OTEL logs. |

132| `otel.trace_exporter.<id>.headers` | `map<string,string>` | Static headers included with OTEL trace exporter requests. |159| `otel.trace_exporter.<id>.headers` | `map<string,string>` | Static headers included with OTEL trace exporter requests. |

134| `otel.trace_exporter.<id>.tls.ca-certificate` | `string` | CA certificate path for OTEL trace exporter TLS. |161| `otel.trace_exporter.<id>.tls.ca-certificate` | `string` | CA certificate path for OTEL trace exporter TLS. |

135| `otel.trace_exporter.<id>.tls.client-certificate` | `string` | Client certificate path for OTEL trace exporter TLS. |162| `otel.trace_exporter.<id>.tls.client-certificate` | `string` | Client certificate path for OTEL trace exporter TLS. |

136| `otel.trace_exporter.<id>.tls.client-private-key` | `string` | Client private key path for OTEL trace exporter TLS. |163| `otel.trace_exporter.<id>.tls.client-private-key` | `string` | Client private key path for OTEL trace exporter TLS. |

164| `permissions.<name>.filesystem` | `table` | Named filesystem permission profile. Each key is an absolute path or special token such as `:minimal` or `:project_roots`. |

165| `permissions.<name>.filesystem.":project_roots".<subpath-or-glob>` | `"read" | "write" | "none"` | Scoped filesystem access relative to the detected project roots. Use `"."` for the root itself; glob subpaths such as `"**/*.env"` can deny reads with `"none"`. |

166| `permissions.<name>.filesystem.<path-or-glob>` | `"read" | "write" | "none" | table` | Grant direct access for a path, glob pattern, or special token, or scope nested entries under that root. Use `"none"` to deny reads for matching paths. |

167| `permissions.<name>.filesystem.glob_scan_max_depth` | `number` | Maximum depth for expanding deny-read glob patterns on platforms that snapshot matches before sandbox startup. Must be at least `1` when set. |

168| `permissions.<name>.network.allow_local_binding` | `boolean` | Permit local bind/listen operations through the managed proxy. |

169| `permissions.<name>.network.allow_upstream_proxy` | `boolean` | Allow the managed proxy to chain to another upstream proxy. |

170| `permissions.<name>.network.dangerously_allow_all_unix_sockets` | `boolean` | Allow the proxy to use arbitrary Unix sockets instead of the default restricted set. |

171| `permissions.<name>.network.dangerously_allow_non_loopback_proxy` | `boolean` | Permit non-loopback bind addresses for the managed proxy listener. |

172| `permissions.<name>.network.domains` | `map<string, allow | deny>` | Domain rules for the managed proxy. Use domain names or wildcard patterns as keys, with `allow` or `deny` values. |

173| `permissions.<name>.network.enable_socks5` | `boolean` | Expose a SOCKS5 listener when this permissions profile enables the managed network proxy. |

174| `permissions.<name>.network.enable_socks5_udp` | `boolean` | Allow UDP over the SOCKS5 listener when enabled. |

175| `permissions.<name>.network.enabled` | `boolean` | Enable network access for this named permissions profile. |

176| `permissions.<name>.network.mode` | `limited | full` | Network proxy mode used for subprocess traffic. |

177| `permissions.<name>.network.proxy_url` | `string` | HTTP proxy endpoint used when this permissions profile enables the managed network proxy. |

178| `permissions.<name>.network.socks_url` | `string` | SOCKS5 proxy endpoint used by this permissions profile. |

179| `permissions.<name>.network.unix_sockets` | `map<string, allow | none>` | Unix socket rules for the managed proxy. Use socket paths as keys, with `allow` or `none` values. |

137| `personality` | `none | friendly | pragmatic` | Default communication style for models that advertise `supportsPersonality`; can be overridden per thread/turn or via `/personality`. |180| `personality` | `none | friendly | pragmatic` | Default communication style for models that advertise `supportsPersonality`; can be overridden per thread/turn or via `/personality`. |

181| `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. |

138| `profile` | `string` | Default profile applied at startup (equivalent to `--profile`). |182| `profile` | `string` | Default profile applied at startup (equivalent to `--profile`). |

139| `profiles.<name>.*` | `various` | Profile-scoped overrides for any of the supported configuration keys. |183| `profiles.<name>.*` | `various` | Profile-scoped overrides for any of the supported configuration keys. |

140| `profiles.<name>.experimental_use_freeform_apply_patch` | `boolean` | Legacy name for enabling freeform apply\_patch; prefer `[features].apply_patch_freeform`. |184| `profiles.<name>.analytics.enabled` | `boolean` | Profile-scoped analytics enablement override. |

141| `profiles.<name>.experimental_use_unified_exec_tool` | `boolean` | Legacy name for enabling unified exec; prefer `[features].unified_exec`. |185| `profiles.<name>.experimental_use_unified_exec_tool` | `boolean` | Legacy name for enabling unified exec; prefer `[features].unified_exec`. |

142| `profiles.<name>.include_apply_patch_tool` | `boolean` | Legacy name for enabling freeform apply\_patch; prefer `[features].apply_patch_freeform`. |

143| `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). |186| `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). |

187| `profiles.<name>.model_instructions_file` | `string (path)` | Profile-scoped replacement for the built-in instruction file. |

144| `profiles.<name>.oss_provider` | `lmstudio | ollama` | Profile-scoped OSS provider for `--oss` sessions. |188| `profiles.<name>.oss_provider` | `lmstudio | ollama` | Profile-scoped OSS provider for `--oss` sessions. |

145| `profiles.<name>.personality` | `none | friendly | pragmatic` | Profile-scoped communication style override for supported models. |189| `profiles.<name>.personality` | `none | friendly | pragmatic` | Profile-scoped communication style override for supported models. |

190| `profiles.<name>.plan_mode_reasoning_effort` | `none | minimal | low | medium | high | xhigh` | Profile-scoped Plan-mode reasoning override. |

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

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

146| `profiles.<name>.web_search` | `disabled | cached | live` | Profile-scoped web search mode override (default: `"cached"`). |193| `profiles.<name>.web_search` | `disabled | cached | live` | Profile-scoped web search mode override (default: `"cached"`). |

194| `profiles.<name>.windows.sandbox` | `unelevated | elevated` | Profile-scoped Windows sandbox mode override. |

147| `project_doc_fallback_filenames` | `array<string>` | Additional filenames to try when `AGENTS.md` is missing. |195| `project_doc_fallback_filenames` | `array<string>` | Additional filenames to try when `AGENTS.md` is missing. |

148| `project_doc_max_bytes` | `number` | Maximum bytes read from `AGENTS.md` when building project instructions. |196| `project_doc_max_bytes` | `number` | Maximum bytes read from `AGENTS.md` when building project instructions. |

149| `project_root_markers` | `array<string>` | List of project root marker filenames; used when searching parent directories for the project root. |197| `project_root_markers` | `array<string>` | List of project root marker filenames; used when searching parent directories for the project root. |

154| `sandbox_workspace_write.exclude_tmpdir_env_var` | `boolean` | Exclude `$TMPDIR` from writable roots in workspace-write mode. |202| `sandbox_workspace_write.exclude_tmpdir_env_var` | `boolean` | Exclude `$TMPDIR` from writable roots in workspace-write mode. |

155| `sandbox_workspace_write.network_access` | `boolean` | Allow outbound network access inside the workspace-write sandbox. |203| `sandbox_workspace_write.network_access` | `boolean` | Allow outbound network access inside the workspace-write sandbox. |

156| `sandbox_workspace_write.writable_roots` | `array<string>` | Additional writable roots when `sandbox_mode = "workspace-write"`. |204| `sandbox_workspace_write.writable_roots` | `array<string>` | Additional writable roots when `sandbox_mode = "workspace-write"`. |

205| `service_tier` | `flex | fast` | Preferred service tier for new turns. |

157| `shell_environment_policy.exclude` | `array<string>` | Glob patterns for removing environment variables after the defaults. |206| `shell_environment_policy.exclude` | `array<string>` | Glob patterns for removing environment variables after the defaults. |

158| `shell_environment_policy.experimental_use_profile` | `boolean` | Use the user shell profile when spawning subprocesses. |207| `shell_environment_policy.experimental_use_profile` | `boolean` | Use the user shell profile when spawning subprocesses. |

159| `shell_environment_policy.ignore_default_excludes` | `boolean` | Keep variables containing KEY/SECRET/TOKEN before other filters run. |208| `shell_environment_policy.ignore_default_excludes` | `boolean` | Keep variables containing KEY/SECRET/TOKEN before other filters run. |

164| `skills.config` | `array<object>` | Per-skill enablement overrides stored in config.toml. |213| `skills.config` | `array<object>` | Per-skill enablement overrides stored in config.toml. |

165| `skills.config.<index>.enabled` | `boolean` | Enable or disable the referenced skill. |214| `skills.config.<index>.enabled` | `boolean` | Enable or disable the referenced skill. |

166| `skills.config.<index>.path` | `string (path)` | Path to a skill folder containing `SKILL.md`. |215| `skills.config.<index>.path` | `string (path)` | Path to a skill folder containing `SKILL.md`. |

216| `sqlite_home` | `string (path)` | Directory where Codex stores the SQLite-backed state DB used by agent jobs and other resumable runtime state. |

167| `suppress_unstable_features_warning` | `boolean` | Suppress the warning that appears when under-development feature flags are enabled. |217| `suppress_unstable_features_warning` | `boolean` | Suppress the warning that appears when under-development feature flags are enabled. |

168| `tool_output_token_limit` | `number` | Token budget for storing individual tool/function outputs in history. |218| `tool_output_token_limit` | `number` | Token budget for storing individual tool/function outputs in history. |

169| `tools.web_search` | `boolean` | Deprecated legacy toggle for web search; prefer the top-level `web_search` setting. |219| `tool_suggest.discoverables` | `array<table>` | Allow tool suggestions for additional discoverable connectors or plugins. Each entry uses `type = "connector"` or `"plugin"` and an `id`. |

220| `tools.view_image` | `boolean` | Enable the local-image attachment tool `view_image`. |

221| `tools.web_search` | `boolean | { context_size = "low|medium|high", allowed_domains = [string], location = { country, region, city, timezone } }` | Optional web search tool configuration. The legacy boolean form is still accepted, but the object form lets you set search context size, allowed domains, and approximate user location. |

170| `tui` | `table` | TUI-specific options such as enabling inline desktop notifications. |222| `tui` | `table` | TUI-specific options such as enabling inline desktop notifications. |

171| `tui.alternate_screen` | `auto | always | never` | Control alternate screen usage for the TUI (default: auto; auto skips it in Zellij to preserve scrollback). |223| `tui.alternate_screen` | `auto | always | never` | Control alternate screen usage for the TUI (default: auto; auto skips it in Zellij to preserve scrollback). |

172| `tui.animations` | `boolean` | Enable terminal animations (welcome screen, shimmer, spinner) (default: true). |224| `tui.animations` | `boolean` | Enable terminal animations (welcome screen, shimmer, spinner) (default: true). |

173| `tui.notification_method` | `auto | osc9 | bel` | Notification method for unfocused terminal notifications (default: auto). |225| `tui.model_availability_nux.<model>` | `integer` | Internal startup-tooltip state keyed by model slug. |

226| `tui.notification_condition` | `unfocused | always` | Control whether TUI notifications fire only when the terminal is unfocused or regardless of focus. Defaults to `unfocused`. |

227| `tui.notification_method` | `auto | osc9 | bel` | Notification method for terminal notifications (default: auto). |

174| `tui.notifications` | `boolean | array<string>` | Enable TUI notifications; optionally restrict to specific event types. |228| `tui.notifications` | `boolean | array<string>` | Enable TUI notifications; optionally restrict to specific event types. |

175| `tui.show_tooltips` | `boolean` | Show onboarding tooltips in the TUI welcome screen (default: true). |229| `tui.show_tooltips` | `boolean` | Show onboarding tooltips in the TUI welcome screen (default: true). |

176| `tui.status_line` | `array<string> | null` | Ordered list of TUI footer status-line item identifiers. `null` disables the status line. |230| `tui.status_line` | `array<string> | null` | Ordered list of TUI footer status-line item identifiers. `null` disables the status line. |

231| `tui.terminal_title` | `array<string> | null` | Ordered list of terminal window/tab title item identifiers. Defaults to `["spinner", "project"]`; `null` disables title updates. |

232| `tui.theme` | `string` | Syntax-highlighting theme override (kebab-case theme name). |

177| `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. |233| `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. |

178| `windows_wsl_setup_acknowledged` | `boolean` | Track Windows onboarding acknowledgement (Windows only). |234| `windows_wsl_setup_acknowledged` | `boolean` | Track Windows onboarding acknowledgement (Windows only). |

179| `windows.sandbox` | `unelevated | elevated` | Windows-only native sandbox mode when running Codex natively on Windows. |235| `windows.sandbox` | `unelevated | elevated` | Windows-only native sandbox mode when running Codex natively on Windows. |

236| `windows.sandbox_private_desktop` | `boolean` | 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. |

180 237 

181Key238Key

182 239 

204 261 

205Key262Key

206 263 

264`agents.<name>.nickname_candidates`

265 

266Type / Values

267 

268`array<string>`

269 

270Details

271 

272Optional pool of display nicknames for spawned agents in that role.

273 

274Key

275 

276`agents.job_max_runtime_seconds`

277 

278Type / Values

279 

280`number`

281 

282Details

283 

284Default per-worker timeout for `spawn_agents_on_csv` jobs. When unset, the tool falls back to 1800 seconds per worker.

285 

286Key

287 

207`agents.max_depth`288`agents.max_depth`

208 289 

209Type / Values290Type / Values

224 305 

225Details306Details

226 307 

227Maximum number of agent threads that can be open concurrently.308Maximum number of agent threads that can be open concurrently. Defaults to `6` when unset.

228 309 

229Key310Key

230 311 

240 321 

241Key322Key

242 323 

324`analytics.enabled`

325 

326Type / Values

327 

328`boolean`

329 

330Details

331 

332Enable or disable analytics for this machine/profile. When unset, the client default applies.

333 

334Key

335 

243`approval_policy`336`approval_policy`

244 337 

245Type / Values338Type / Values

246 339 

247`untrusted | on-request | never | { reject = { sandbox_approval = bool, rules = bool, mcp_elicitations = bool } }`340`untrusted | on-request | never | { granular = { sandbox_approval = bool, rules = bool, mcp_elicitations = bool, request_permissions = bool, skill_approval = bool } }`

341 

342Details

343 

344Controls 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.

345 

346Key

347 

348`approval_policy.granular.mcp_elicitations`

349 

350Type / Values

351 

352`boolean`

353 

354Details

355 

356When `true`, MCP elicitation prompts are allowed to surface instead of being auto-rejected.

357 

358Key

359 

360`approval_policy.granular.request_permissions`

361 

362Type / Values

363 

364`boolean`

248 365 

249Details366Details

250 367 

251Controls 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.368When `true`, prompts from the `request_permissions` tool are allowed to surface.

252 369 

253Key370Key

254 371 

255`approval_policy.reject.mcp_elicitations`372`approval_policy.granular.rules`

256 373 

257Type / Values374Type / Values

258 375 

260 377 

261Details378Details

262 379 

263When `true`, MCP elicitation prompts are auto-rejected instead of shown to the user.380When `true`, approvals triggered by execpolicy `prompt` rules are allowed to surface.

264 381 

265Key382Key

266 383 

267`approval_policy.reject.rules`384`approval_policy.granular.sandbox_approval`

268 385 

269Type / Values386Type / Values

270 387 

272 389 

273Details390Details

274 391 

275When `true`, approvals triggered by execpolicy `prompt` rules are auto-rejected.392When `true`, sandbox escalation approval prompts are allowed to surface.

276 393 

277Key394Key

278 395 

279`approval_policy.reject.sandbox_approval`396`approval_policy.granular.skill_approval`

280 397 

281Type / Values398Type / Values

282 399 

284 401 

285Details402Details

286 403 

287When `true`, sandbox escalation approval prompts are auto-rejected.404When `true`, skill-script approval prompts are allowed to surface.

405 

406Key

407 

408`approvals_reviewer`

409 

410Type / Values

411 

412`user | guardian_subagent`

413 

414Details

415 

416Select who reviews eligible approval prompts. Defaults to `user`; `guardian_subagent` routes supported reviews through the Guardian reviewer subagent.

288 417 

289Key418Key

290 419 

456 585 

457Key586Key

458 587 

459`compact_prompt`588`commit_attribution`

460 589 

461Type / Values590Type / Values

462 591 

464 593 

465Details594Details

466 595 

467Inline override for the history compaction prompt.596Override the commit co-author trailer text. Set an empty string to disable automatic attribution.

468 597 

469Key598Key

470 599 

471`developer_instructions`600`compact_prompt`

472 601 

473Type / Values602Type / Values

474 603 

476 605 

477Details606Details

478 607 

479Additional developer instructions injected into the session (optional).608Inline override for the history compaction prompt.

480 609 

481Key610Key

482 611 

483`disable_paste_burst`612`default_permissions`

484 613 

485Type / Values614Type / Values

486 615 

487`boolean`616`string`

488 617 

489Details618Details

490 619 

491Disable burst-paste detection in the TUI.620Name of the default permissions profile to apply to sandboxed tool calls.

492 621 

493Key622Key

494 623 

495`experimental_compact_prompt_file`624`developer_instructions`

496 625 

497Type / Values626Type / Values

498 627 

499`string (path)`628`string`

500 629 

501Details630Details

502 631 

503Load the compaction prompt override from a file (experimental).632Additional developer instructions injected into the session (optional).

504 633 

505Key634Key

506 635 

507`experimental_use_freeform_apply_patch`636`disable_paste_burst`

508 637 

509Type / Values638Type / Values

510 639 

512 641 

513Details642Details

514 643 

515Legacy name for enabling freeform apply\_patch; prefer `[features].apply_patch_freeform` or `codex --enable apply_patch_freeform`.644Disable burst-paste detection in the TUI.

516 645 

517Key646Key

518 647 

519`experimental_use_unified_exec_tool`648`experimental_compact_prompt_file`

520 649 

521Type / Values650Type / Values

522 651 

523`boolean`652`string (path)`

524 653 

525Details654Details

526 655 

527Legacy name for enabling unified exec; prefer `[features].unified_exec` or `codex --enable unified_exec`.656Load the compaction prompt override from a file (experimental).

528 657 

529Key658Key

530 659 

531`features.apply_patch_freeform`660`experimental_use_unified_exec_tool`

532 661 

533Type / Values662Type / Values

534 663 

536 665 

537Details666Details

538 667 

539Expose the freeform `apply_patch` tool (experimental).668Legacy name for enabling unified exec; prefer `[features].unified_exec` or `codex --enable unified_exec`.

540 669 

541Key670Key

542 671 

552 681 

553Key682Key

554 683 

555`features.apps_mcp_gateway`684`features.codex_hooks`

556 685 

557Type / Values686Type / Values

558 687 

560 689 

561Details690Details

562 691 

563Route Apps MCP calls through the OpenAI connectors MCP gateway (`https://api.openai.com/v1/connectors/mcp/`) instead of legacy routing (experimental).692Enable lifecycle hooks loaded from `hooks.json` (under development; off by default).

564 693 

565Key694Key

566 695 

567`features.child_agents_md`696`features.enable_request_compression`

568 697 

569Type / Values698Type / Values

570 699 

572 701 

573Details702Details

574 703 

575Append AGENTS.md scope/precedence guidance even when no AGENTS.md is present (experimental).704Compress streaming request bodies with zstd when supported (stable; on by default).

576 705 

577Key706Key

578 707 

579`features.collaboration_modes`708`features.fast_mode`

580 709 

581Type / Values710Type / Values

582 711 

584 713 

585Details714Details

586 715 

587Enable collaboration modes such as plan mode (stable; on by default).716Enable Fast mode selection and the `service_tier = "fast"` path (stable; on by default).

588 717 

589Key718Key

590 719 

591`features.multi_agent`720`features.guardian_approval`

592 721 

593Type / Values722Type / Values

594 723 

596 725 

597Details726Details

598 727 

599Enable multi-agent collaboration tools (`spawn\_agent`, `send\_input`, `resume\_agent`, `wait`, and `close\_agent`) (experimental; off by default).728Route eligible approval requests through the guardian reviewer subagent (experimental; off by default). Use with `approvals_reviewer = "guardian_subagent"`.

600 729 

601Key730Key

602 731 

603`features.personality`732`features.memories`

604 733 

605Type / Values734Type / Values

606 735 

608 737 

609Details738Details

610 739 

611Enable personality selection controls (stable; on by default).740Enable [Memories](https://developers.openai.com/codex/memories) (off by default).

612 741 

613Key742Key

614 743 

615`features.powershell_utf8`744`features.multi_agent`

616 745 

617Type / Values746Type / Values

618 747 

620 749 

621Details750Details

622 751 

623Force PowerShell UTF-8 output (defaults to true).752Enable multi-agent collaboration tools (`spawn_agent`, `send_input`, `resume_agent`, `wait_agent`, and `close_agent`) (stable; on by default).

624 753 

625Key754Key

626 755 

627`features.remote_models`756`features.personality`

628 757 

629Type / Values758Type / Values

630 759 

632 761 

633Details762Details

634 763 

635Refresh remote model list before showing readiness (experimental).764Enable personality selection controls (stable; on by default).

636 765 

637Key766Key

638 767 

639`features.request_rule`768`features.prevent_idle_sleep`

640 769 

641Type / Values770Type / Values

642 771 

644 773 

645Details774Details

646 775 

647Enable Smart approvals (`prefix_rule` suggestions on escalation requests; stable; on by default).776Prevent the machine from sleeping while a turn is actively running (experimental; off by default).

648 777 

649Key778Key

650 779 

651`features.runtime_metrics`780`features.shell_snapshot`

652 781 

653Type / Values782Type / Values

654 783 

656 785 

657Details786Details

658 787 

659Show runtime metrics summary in TUI turn separators (experimental).788Snapshot shell environment to speed up repeated commands (stable; on by default).

660 789 

661Key790Key

662 791 

663`features.search_tool`792`features.shell_tool`

664 793 

665Type / Values794Type / Values

666 795 

668 797 

669Details798Details

670 799 

671Enable `search_tool_bm25` for Apps tool discovery before invoking app MCP tools (experimental).800Enable the default `shell` tool for running commands (stable; on by default).

672 801 

673Key802Key

674 803 

675`features.shell_snapshot`804`features.skill_mcp_dependency_install`

676 805 

677Type / Values806Type / Values

678 807 

680 809 

681Details810Details

682 811 

683Snapshot shell environment to speed up repeated commands (beta).812Allow prompting and installing missing MCP dependencies for skills (stable; on by default).

684 813 

685Key814Key

686 815 

687`features.shell_tool`816`features.undo`

688 817 

689Type / Values818Type / Values

690 819 

692 821 

693Details822Details

694 823 

695Enable the default `shell` tool for running commands (stable; on by default).824Enable undo support (stable; off by default).

696 825 

697Key826Key

698 827 

704 833 

705Details834Details

706 835 

707Use the unified PTY-backed exec tool (beta).836Use the unified PTY-backed exec tool (stable; enabled by default except on Windows).

708 

709Key

710 

711`features.use_linux_sandbox_bwrap`

712 

713Type / Values

714 

715`boolean`

716 

717Details

718 

719Use the bubblewrap-based Linux sandbox pipeline (experimental; off by default).

720 837 

721Key838Key

722 839 

840 957 

841Key958Key

842 959 

843`include_apply_patch_tool`

844 

845Type / Values

846 

847`boolean`

848 

849Details

850 

851Legacy name for enabling freeform apply\_patch; prefer `[features].apply_patch_freeform`.

852 

853Key

854 

855`instructions`960`instructions`

856 961 

857Type / Values962Type / Values

1024 1129 

1025Type / Values1130Type / Values

1026 1131 

1027`array<string>`1132`array<string | { name = string, source = "local" | "remote" }>`

1133 

1134Details

1135 

1136Additional environment variables to whitelist for an MCP stdio server. String entries default to `source = "local"`; use `source = "remote"` only with executor-backed remote stdio.

1137 

1138Key

1139 

1140`mcp_servers.<id>.experimental_environment`

1141 

1142Type / Values

1143 

1144`local | remote`

1028 1145 

1029Details1146Details

1030 1147 

1031Additional environment variables to whitelist for an MCP stdio server.1148Experimental placement for an MCP server. `remote` starts stdio servers through a remote executor environment; streamable HTTP remote placement is not implemented.

1032 1149 

1033Key1150Key

1034 1151 

1044 1161 

1045Key1162Key

1046 1163 

1164`mcp_servers.<id>.oauth_resource`

1165 

1166Type / Values

1167 

1168`string`

1169 

1170Details

1171 

1172Optional RFC 8707 OAuth resource parameter to include during MCP login.

1173 

1174Key

1175 

1047`mcp_servers.<id>.required`1176`mcp_servers.<id>.required`

1048 1177 

1049Type / Values1178Type / Values

1056 1185 

1057Key1186Key

1058 1187 

1188`mcp_servers.<id>.scopes`

1189 

1190Type / Values

1191 

1192`array<string>`

1193 

1194Details

1195 

1196OAuth scopes to request when authenticating to that MCP server.

1197 

1198Key

1199 

1059`mcp_servers.<id>.startup_timeout_ms`1200`mcp_servers.<id>.startup_timeout_ms`

1060 1201 

1061Type / Values1202Type / Values

1104 1245 

1105Key1246Key

1106 1247 

1107`model`1248`memories.consolidation_model`

1108 1249 

1109Type / Values1250Type / Values

1110 1251 

1112 1253 

1113Details1254Details

1114 1255 

1115Model to use (e.g., `gpt-5-codex`).1256Optional model override for global memory consolidation.

1116 1257 

1117Key1258Key

1118 1259 

1119`model_auto_compact_token_limit`1260`memories.disable_on_external_context`

1120 1261 

1121Type / Values1262Type / Values

1122 1263 

1123`number`1264`boolean`

1124 1265 

1125Details1266Details

1126 1267 

1127Token threshold that triggers automatic history compaction (unset uses model defaults).1268When `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`.

1128 1269 

1129Key1270Key

1130 1271 

1131`model_catalog_json`1272`memories.extract_model`

1132 1273 

1133Type / Values1274Type / Values

1134 1275 

1135`string (path)`1276`string`

1136 1277 

1137Details1278Details

1138 1279 

1139Optional path to a JSON model catalog loaded on startup. Profile-level `profiles.<name>.model_catalog_json` can override this per profile.1280Optional model override for per-thread memory extraction.

1140 1281 

1141Key1282Key

1142 1283 

1143`model_context_window`1284`memories.generate_memories`

1144 1285 

1145Type / Values1286Type / Values

1146 1287 

1147`number`1288`boolean`

1148 1289 

1149Details1290Details

1150 1291 

1151Context window tokens available to the active model.1292When `false`, newly created threads are not stored as memory-generation inputs. Defaults to `true`.

1152 1293 

1153Key1294Key

1154 1295 

1155`model_instructions_file`1296`memories.max_raw_memories_for_consolidation`

1156 1297 

1157Type / Values1298Type / Values

1158 1299 

1159`string (path)`1300`number`

1160 1301 

1161Details1302Details

1162 1303 

1163Replacement for built-in instructions instead of `AGENTS.md`.1304Maximum recent raw memories retained for global consolidation. Defaults to `256` and is capped at `4096`.

1164 1305 

1165Key1306Key

1166 1307 

1167`model_provider`1308`memories.max_rollout_age_days`

1168 1309 

1169Type / Values1310Type / Values

1170 1311 

1171`string`1312`number`

1172 1313 

1173Details1314Details

1174 1315 

1175Provider id from `model_providers` (default: `openai`).1316Maximum age of threads considered for memory generation. Defaults to `30` and is clamped to `0`-`90`.

1176 1317 

1177Key1318Key

1178 1319 

1179`model_providers.<id>.base_url`1320`memories.max_rollouts_per_startup`

1180 1321 

1181Type / Values1322Type / Values

1182 1323 

1183`string`1324`number`

1184 1325 

1185Details1326Details

1186 1327 

1187API base URL for the model provider.1328Maximum rollout candidates processed per startup pass. Defaults to `16` and is capped at `128`.

1188 1329 

1189Key1330Key

1190 1331 

1191`model_providers.<id>.env_http_headers`1332`memories.max_unused_days`

1192 1333 

1193Type / Values1334Type / Values

1194 1335 

1195`map<string,string>`1336`number`

1196 1337 

1197Details1338Details

1198 1339 

1199HTTP headers populated from environment variables when present.1340Maximum days since a memory was last used before it becomes ineligible for consolidation. Defaults to `30` and is clamped to `0`-`365`.

1200 1341 

1201Key1342Key

1202 1343 

1203`model_providers.<id>.env_key`1344`memories.min_rollout_idle_hours`

1204 1345 

1205Type / Values1346Type / Values

1206 1347 

1207`string`1348`number`

1208 1349 

1209Details1350Details

1210 1351 

1211Environment variable supplying the provider API key.1352Minimum idle time before a thread is considered for memory generation. Defaults to `6` and is clamped to `1`-`48`.

1212 1353 

1213Key1354Key

1214 1355 

1215`model_providers.<id>.env_key_instructions`1356`memories.use_memories`

1216 1357 

1217Type / Values1358Type / Values

1218 1359 

1219`string`1360`boolean`

1220 1361 

1221Details1362Details

1222 1363 

1223Optional setup guidance for the provider API key.1364When `false`, Codex skips injecting existing memories into future sessions. Defaults to `true`.

1224 1365 

1225Key1366Key

1226 1367 

1227`model_providers.<id>.experimental_bearer_token`1368`model`

1228 1369 

1229Type / Values1370Type / Values

1230 1371 

1232 1373 

1233Details1374Details

1234 1375 

1235Direct bearer token for the provider (discouraged; use `env_key`).1376Model to use (e.g., `gpt-5.4`).

1236 1377 

1237Key1378Key

1238 1379 

1239`model_providers.<id>.http_headers`1380`model_auto_compact_token_limit`

1240 1381 

1241Type / Values1382Type / Values

1242 1383 

1243`map<string,string>`1384`number`

1244 1385 

1245Details1386Details

1246 1387 

1247Static HTTP headers added to provider requests.1388Token threshold that triggers automatic history compaction (unset uses model defaults).

1248 1389 

1249Key1390Key

1250 1391 

1251`model_providers.<id>.name`1392`model_catalog_json`

1252 1393 

1253Type / Values1394Type / Values

1254 1395 

1255`string`1396`string (path)`

1256 1397 

1257Details1398Details

1258 1399 

1259Display name for a custom model provider.1400Optional path to a JSON model catalog loaded on startup. Profile-level `profiles.<name>.model_catalog_json` can override this per profile.

1260 1401 

1261Key1402Key

1262 1403 

1263`model_providers.<id>.query_params`1404`model_context_window`

1264 1405 

1265Type / Values1406Type / Values

1266 1407 

1267`map<string,string>`1408`number`

1268 1409 

1269Details1410Details

1270 1411 

1271Extra query parameters appended to provider requests.1412Context window tokens available to the active model.

1272 1413 

1273Key1414Key

1274 1415 

1275`model_providers.<id>.request_max_retries`1416`model_instructions_file`

1276 1417 

1277Type / Values1418Type / Values

1278 1419 

1279`number`1420`string (path)`

1280 1421 

1281Details1422Details

1282 1423 

1283Retry count for HTTP requests to the provider (default: 4).1424Replacement for built-in instructions instead of `AGENTS.md`.

1284 1425 

1285Key1426Key

1286 1427 

1287`model_providers.<id>.requires_openai_auth`1428`model_provider`

1288 1429 

1289Type / Values1430Type / Values

1290 1431 

1291`boolean`1432`string`

1292 1433 

1293Details1434Details

1294 1435 

1295The provider uses OpenAI authentication (defaults to false).1436Provider id from `model_providers` (default: `openai`).

1296 1437 

1297Key1438Key

1298 1439 

1299`model_providers.<id>.stream_idle_timeout_ms`1440`model_providers.<id>`

1300 1441 

1301Type / Values1442Type / Values

1302 1443 

1303`number`1444`table`

1304 1445 

1305Details1446Details

1306 1447 

1307Idle timeout for SSE streams in milliseconds (default: 300000).1448Custom provider definition. Built-in provider IDs (`openai`, `ollama`, and `lmstudio`) are reserved and cannot be overridden.

1308 1449 

1309Key1450Key

1310 1451 

1311`model_providers.<id>.stream_max_retries`1452`model_providers.<id>.auth`

1312 1453 

1313Type / Values1454Type / Values

1314 1455 

1315`number`1456`table`

1316 1457 

1317Details1458Details

1318 1459 

1319Retry count for SSE streaming interruptions (default: 5).1460Command-backed bearer token configuration for a custom provider. Do not combine with `env_key`, `experimental_bearer_token`, or `requires_openai_auth`.

1320 1461 

1321Key1462Key

1322 1463 

1323`model_providers.<id>.wire_api`1464`model_providers.<id>.auth.args`

1324 1465 

1325Type / Values1466Type / Values

1326 1467 

1327`chat | responses`1468`array<string>`

1328 1469 

1329Details1470Details

1330 1471 

1331Protocol used by the provider (defaults to `chat` if omitted).1472Arguments passed to the token command.

1332 1473 

1333Key1474Key

1334 1475 

1335`model_reasoning_effort`1476`model_providers.<id>.auth.command`

1336 1477 

1337Type / Values1478Type / Values

1338 1479 

1339`minimal | low | medium | high | xhigh`1480`string`

1340 1481 

1341Details1482Details

1342 1483 

1343Adjust reasoning effort for supported models (Responses API only; `xhigh` is model-dependent).1484Command to run when Codex needs a bearer token. The command must print the token to stdout.

1344 1485 

1345Key1486Key

1346 1487 

1347`model_reasoning_summary`1488`model_providers.<id>.auth.cwd`

1348 1489 

1349Type / Values1490Type / Values

1350 1491 

1351`auto | concise | detailed | none`1492`string (path)`

1352 1493 

1353Details1494Details

1354 1495 

1355Select reasoning summary detail or disable summaries entirely.1496Working directory for the token command.

1356 1497 

1357Key1498Key

1358 1499 

1359`model_supports_reasoning_summaries`1500`model_providers.<id>.auth.refresh_interval_ms`

1360 1501 

1361Type / Values1502Type / Values

1362 1503 

1363`boolean`1504`number`

1364 1505 

1365Details1506Details

1366 1507 

1367Force Codex to send or not send reasoning metadata.1508How often Codex proactively refreshes the token in milliseconds (default: 300000). Set to `0` to refresh only after an authentication retry.

1368 1509 

1369Key1510Key

1370 1511 

1371`model_verbosity`1512`model_providers.<id>.auth.timeout_ms`

1372 1513 

1373Type / Values1514Type / Values

1374 1515 

1375`low | medium | high`1516`number`

1376 1517 

1377Details1518Details

1378 1519 

1379Control GPT-5 Responses API verbosity (defaults to `medium`).1520Maximum token command runtime in milliseconds (default: 5000).

1380 1521 

1381Key1522Key

1382 1523 

1383`notice.hide_full_access_warning`1524`model_providers.<id>.base_url`

1384 1525 

1385Type / Values1526Type / Values

1386 1527 

1387`boolean`1528`string`

1388 1529 

1389Details1530Details

1390 1531 

1391Track acknowledgement of the full access warning prompt.1532API base URL for the model provider.

1392 1533 

1393Key1534Key

1394 1535 

1395`notice.hide_gpt-5.1-codex-max_migration_prompt`1536`model_providers.<id>.env_http_headers`

1396 1537 

1397Type / Values1538Type / Values

1398 1539 

1399`boolean`1540`map<string,string>`

1400 1541 

1401Details1542Details

1402 1543 

1403Track acknowledgement of the gpt-5.1-codex-max migration prompt.1544HTTP headers populated from environment variables when present.

1404 1545 

1405Key1546Key

1406 1547 

1407`notice.hide_gpt5_1_migration_prompt`1548`model_providers.<id>.env_key`

1549 

1550Type / Values

1551 

1552`string`

1553 

1554Details

1555 

1556Environment variable supplying the provider API key.

1557 

1558Key

1559 

1560`model_providers.<id>.env_key_instructions`

1561 

1562Type / Values

1563 

1564`string`

1565 

1566Details

1567 

1568Optional setup guidance for the provider API key.

1569 

1570Key

1571 

1572`model_providers.<id>.experimental_bearer_token`

1573 

1574Type / Values

1575 

1576`string`

1577 

1578Details

1579 

1580Direct bearer token for the provider (discouraged; use `env_key`).

1581 

1582Key

1583 

1584`model_providers.<id>.http_headers`

1585 

1586Type / Values

1587 

1588`map<string,string>`

1589 

1590Details

1591 

1592Static HTTP headers added to provider requests.

1593 

1594Key

1595 

1596`model_providers.<id>.name`

1597 

1598Type / Values

1599 

1600`string`

1601 

1602Details

1603 

1604Display name for a custom model provider.

1605 

1606Key

1607 

1608`model_providers.<id>.query_params`

1609 

1610Type / Values

1611 

1612`map<string,string>`

1613 

1614Details

1615 

1616Extra query parameters appended to provider requests.

1617 

1618Key

1619 

1620`model_providers.<id>.request_max_retries`

1621 

1622Type / Values

1623 

1624`number`

1625 

1626Details

1627 

1628Retry count for HTTP requests to the provider (default: 4).

1629 

1630Key

1631 

1632`model_providers.<id>.requires_openai_auth`

1633 

1634Type / Values

1635 

1636`boolean`

1637 

1638Details

1639 

1640The provider uses OpenAI authentication (defaults to false).

1641 

1642Key

1643 

1644`model_providers.<id>.stream_idle_timeout_ms`

1645 

1646Type / Values

1647 

1648`number`

1649 

1650Details

1651 

1652Idle timeout for SSE streams in milliseconds (default: 300000).

1653 

1654Key

1655 

1656`model_providers.<id>.stream_max_retries`

1657 

1658Type / Values

1659 

1660`number`

1661 

1662Details

1663 

1664Retry count for SSE streaming interruptions (default: 5).

1665 

1666Key

1667 

1668`model_providers.<id>.supports_websockets`

1669 

1670Type / Values

1671 

1672`boolean`

1673 

1674Details

1675 

1676Whether that provider supports the Responses API WebSocket transport.

1677 

1678Key

1679 

1680`model_providers.<id>.wire_api`

1681 

1682Type / Values

1683 

1684`responses`

1685 

1686Details

1687 

1688Protocol used by the provider. `responses` is the only supported value, and it is the default when omitted.

1689 

1690Key

1691 

1692`model_reasoning_effort`

1693 

1694Type / Values

1695 

1696`minimal | low | medium | high | xhigh`

1697 

1698Details

1699 

1700Adjust reasoning effort for supported models (Responses API only; `xhigh` is model-dependent).

1701 

1702Key

1703 

1704`model_reasoning_summary`

1705 

1706Type / Values

1707 

1708`auto | concise | detailed | none`

1709 

1710Details

1711 

1712Select reasoning summary detail or disable summaries entirely.

1713 

1714Key

1715