# `ReqManagedAgents.SessionResult`
[🔗](https://github.com/cash-mckeeman/req_managed_agents/blob/v0.16.0/req_managed_agents/lib/req_managed_agents/session_result.ex#L1)

The accumulated outcome of a whole run — what `Session.run/2` and `message/2` deliver.

`transcript` is the provider's client-held history when the provider exports
`transcript/1` (potentially large); `nil` for server-held providers.

`in_flight: :unknown` means additional provider work may be unmeasured;
`:none` marks a completed local boundary. `usage.complete?` describes the
represented contributions, so neither flag alone certifies total billed cost.

`outcomes` is the provider's native grading evidence projected from `events`
(`[]` when the provider has none). It resets with `Session.message/2` and
accumulates across `Session.send_event/2` chained definitions, in
first-observation order, so `hd/1` is not the current outcome: select by
`outcome_id`. A grade is not a completion signal; read `terminal` for that.

# `t`

```elixir
@type t() :: %ReqManagedAgents.SessionResult{
  custom_tool_uses: [ReqManagedAgents.ToolUse.t()],
  events: [map()],
  in_flight: :none | :unknown,
  outcomes: [ReqManagedAgents.OutcomeResult.t()],
  server_tool_uses: [ReqManagedAgents.ToolUse.t()],
  session_id: String.t() | nil,
  stop_reason: String.t() | map() | atom() | nil,
  terminal: ReqManagedAgents.Provider.terminal(),
  text: String.t(),
  transcript: [map()] | nil,
  turns: non_neg_integer(),
  usage: ReqManagedAgents.Usage.t()
}
```

---

*Consult [api-reference.md](api-reference.md) for complete listing*
