Documentation
¶
Index ¶
- Constants
- Variables
- func ContextWindowFor(model string, largeContext bool) int
- func DefaultEffort() string
- func DefaultModel() string
- func SandboxDisabled() bool
- type Agent
- type Config
- type Content
- type File
- type Message
- type MessageRole
- type ModelInfo
- type Reasoning
- type State
- type ToolCall
- type ToolResult
- type Usage
Constants ¶
const ( DefaultMaxTurns = 100 DefaultMaxParallelTools = 8 DefaultToolTimeout = 10 * time.Minute DefaultContextWindow = 400_000 DefaultReserveTokens = 32_000 )
Variables ¶
var ErrEmptyInput = errors.New("agent input is empty")
ErrEmptyInput means Send was called without any content.
var ErrMaxTurnsExceeded = errors.New("agent: internal turn-loop safety bound exceeded — likely a runaway tool-call cycle")
var ErrTurnInProgress = errors.New("agent turn already in progress")
ErrTurnInProgress means Send was called while another turn was active.
Functions ¶
func ContextWindowFor ¶ added in v0.9.3
func DefaultEffort ¶ added in v0.10.5
func DefaultEffort() string
DefaultEffort returns the reasoning effort requested via WINGMAN_EFFORT. Empty (or "auto") leaves the provider default in place. Unrecognized values are ignored so a typo cannot silently pin an unexpected effort.
func DefaultModel ¶ added in v0.9.4
func DefaultModel() string
DefaultModel returns the model requested via environment; WINGMAN_MODEL takes priority over the OpenAI-standard OPENAI_DEFAULT_MODEL.
func SandboxDisabled ¶ added in v0.10.6
func SandboxDisabled() bool
Types ¶
type Agent ¶
type Agent struct {
*Config
Messages []Message
Usage Usage
Revision uint64
// contains filtered or unexported fields
}
func (*Agent) MessagesSnapshot ¶ added in v0.10.3
func (*Agent) QueueInput ¶ added in v0.10.9
QueueInput adds guidance to the active run. The agent consumes queued input at the next safe model boundary. It returns false when no run is active so callers can preserve the input as a normal follow-up instead.
func (*Agent) Send ¶
Send starts exactly one turn. It never queues implicitly: callers that want to guide the active turn must use QueueInput, while FIFO follow-ups belong in a caller-owned session orchestrator. Setup errors are returned immediately; failures after the turn starts are yielded by the returned stream.
func (*Agent) StateSnapshot ¶ added in v0.10.3
func (*Agent) UsageSnapshot ¶ added in v0.10.3
type Config ¶
type Config struct {
Model func() string
Effort func() string
Tools func() []tool.Tool
Instructions func() string
// CacheKey routes provider-side prompt caching; keep it stable per
// conversation (e.g. the session ID) to maximize prefix-cache hits.
CacheKey string
Hooks hook.Hooks
// MaxTurns caps successful model invocations in one Send run. Stream
// retries and tool calls do not consume turns. Zero uses the default;
// negative disables the safety bound.
MaxTurns int
// MaxParallelTools bounds concurrently executing read-only tool calls.
// Zero uses the default; negative allows the whole emitted batch.
MaxParallelTools int
// ToolTimeout is a hard ceiling on every tool call. When zero, tools may
// extend the default via tool.Tool.Timeout; negative disables deadlines.
ToolTimeout time.Duration
ContextWindow int
// LargeContext compacts against the model's full hardware window instead
// of stopping at the provider's long-context price threshold (e.g. 2x
// input pricing on GPT-5.4/5.5 beyond 272k input tokens).
LargeContext bool
ReserveTokens int
// contains filtered or unexported fields
}
func DefaultConfig ¶
type Content ¶
type Content struct {
Text string `json:"text,omitempty"`
Refusal string `json:"refusal,omitempty"`
File *File `json:"file,omitempty"`
Reasoning *Reasoning `json:"reasoning,omitempty"`
ToolCall *ToolCall `json:"tool_call,omitempty"`
ToolResult *ToolResult `json:"tool_result,omitempty"`
}
func CloneContent ¶ added in v0.10.9
CloneContent returns an independent copy suitable for retaining after an API call. Content only contains value fields and one level of pointer fields.
type Message ¶
type Message struct {
Role MessageRole `json:"role"`
Content []Content `json:"content"`
Hidden bool `json:"hidden,omitempty"`
}
type MessageRole ¶
type MessageRole string
const ( RoleUser MessageRole = "user" RoleAssistant MessageRole = "assistant" RoleSystem MessageRole = "system" )
type Reasoning ¶ added in v0.6.2
type Reasoning struct {
ID string `json:"id,omitempty"`
Summary string `json:"summary,omitempty"`
// Content is the provider's opaque (encrypted) reasoning payload, only
// replayable to the model that produced it. Model tags the producer so the
// agent loop can purge stale payloads when the session model changes.
Content string `json:"content,omitempty"`
Model string `json:"model,omitempty"`
}
type State ¶ added in v0.6.2
type ToolResult ¶
type Usage ¶
type Usage struct {
InputTokens int64 `json:"input_tokens"`
CachedTokens int64 `json:"cached_tokens"`
OutputTokens int64 `json:"output_tokens"`
// LastInputTokens is the input size of the most recent request — the
// current context occupancy, unlike the cumulative counters above.
LastInputTokens int64 `json:"last_input_tokens,omitempty"`
}