{ "$schema": "https://json-schema.org/draft/2020-12/schema", "$id": "#/components/schemas/TaskRunBootstrapCreateRequest", "title": "TaskRunBootstrapCreateRequest", "type": "object", "description": "Request body for creating a task run without starting execution yet.", "properties": { "environment": { "allOf": [ { "$ref": "#/components/schemas/TaskRunBootstrapCreateRequestEnvironmentEnum" } ], "default": "local", "description": "Execution environment for the new run. Use 'cloud' for remote sandbox runs and 'local' for desktop sessions.\n\n* `local` - local\n* `cloud` - cloud" }, "mode": { "allOf": [ { "$ref": "#/components/schemas/TaskExecutionModeEnum" } ], "default": "background", "description": "Execution mode: 'interactive' for user-connected runs, 'background' for autonomous runs\n\n* `interactive` - interactive\n* `background` - background" }, "branch": { "type": "string", "nullable": true, "description": "Git branch to checkout in the sandbox", "maxLength": 255 }, "sandbox_environment_id": { "type": "string", "format": "uuid", "description": "Optional sandbox environment to apply for this cloud run." }, "pr_authorship_mode": { "allOf": [ { "$ref": "#/components/schemas/PrAuthorshipModeEnum" } ], "description": "Whether pull requests for this run should be authored by the user or the bot.\n\n* `user` - user\n* `bot` - bot" }, "run_source": { "allOf": [ { "$ref": "#/components/schemas/RunSourceEnum" } ], "description": "High-level source that triggered this run, used to distinguish manual and signal-based cloud runs.\n\n* `manual` - manual\n* `signal_report` - signal_report" }, "signal_report_id": { "type": "string", "description": "Optional signal report identifier when this run was started from Inbox." }, "runtime_adapter": { "allOf": [ { "$ref": "#/components/schemas/RuntimeAdapterEnum" } ], "description": "Agent runtime adapter to launch for this run. Use 'claude' for the Claude runtime or 'codex' for the Codex runtime.\n\n* `claude` - claude\n* `codex` - codex" }, "model": { "type": "string", "description": "LLM model identifier to run in the selected runtime." }, "reasoning_effort": { "allOf": [ { "$ref": "#/components/schemas/ReasoningEffortEnum" } ], "description": "Reasoning effort to request for models that expose an effort control.\n\n* `low` - low\n* `medium` - medium\n* `high` - high\n* `xhigh` - xhigh\n* `max` - max" }, "github_user_token": { "type": "string", "writeOnly": true, "description": "Ephemeral GitHub user token from PostHog Code for user-authored cloud pull requests." }, "initial_permission_mode": { "allOf": [ { "$ref": "#/components/schemas/TaskRunBootstrapCreateRequestInitialPermissionModeEnum" } ], "description": "Initial permission mode for the agent session. Claude runtimes accept PostHog permission presets like 'plan'. Codex runtimes accept native Codex modes like 'auto' and 'read-only'.\n\n* `default` - default\n* `acceptEdits` - acceptEdits\n* `plan` - plan\n* `bypassPermissions` - bypassPermissions\n* `auto` - auto\n* `read-only` - read-only\n* `full-access` - full-access" } } }