ExecutionResult

Agent Runtimes

---

Agent Runtimes / types/execution / ExecutionResult

Interface: ExecutionResult

Defined in: types/execution.ts:34

Result from code or skill execution. This maps to the Python ExecutionResult model from code-sandboxes.

Distinguishes between three levels of outcome:

1. Execution-level failure (execution_ok=false): Infrastructure/sandbox failed
2. Code-level error (code_error present): User code raised Python exception
3. Exit code (exit_code present): User code called sys.exit() with a code

Properties

code_error?

optional code_error?: CodeError | null

Defined in: types/execution.ts:45

Python exception details when code raised an error

---

duration?

optional duration?: number

Defined in: types/execution.ts:60

Execution timing in seconds

---

error?

optional error?: string | null

Defined in: types/execution.ts:57

Backwards-compatible error message (derived from execution_error or code_error)

---

execution_error?

optional execution_error?: string | null

Defined in: types/execution.ts:42

Infrastructure error details when execution_ok is false

---

execution_ok

execution_ok: boolean

Defined in: types/execution.ts:39

Whether the sandbox infrastructure executed the code successfully

---

exit_code?

optional exit_code?: number | null

Defined in: types/execution.ts:48

Exit code when code called sys.exit() - not a failure, but an intentional exit

---

interrupted?

optional interrupted?: boolean

Defined in: types/execution.ts:63

Whether execution was interrupted/cancelled

---

output?

optional output?: string

Defined in: types/execution.ts:54

Standard output text

---

result?

optional result?: unknown[]

Defined in: types/execution.ts:51

Execution results (display outputs, return values)

---

success

success: boolean

Defined in: types/execution.ts:36

Whether overall execution succeeded (no infrastructure or code errors)