Help us shape the future of Go! Take the 2026 Go Developer Survey

webapi

package
v0.39.0 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Oct 20, 2026 License: MIT Imports: 25 Imported by: 0

Documentation

Index

Constants

This section is empty.

Variables

View Source 
var ErrRunNotFound = errors.New("run not found")

ErrRunNotFound is returned when a run ID does not match any stored run.

View Source 
var Version = "0.4.0-alpha.1"

Version is set at build time or defaults to dev.

Functions

func CORSMiddleware 

func CORSMiddleware(next http.Handler, allowedOrigins ...string) http.Handler

CORSMiddleware wraps a handler with CORS headers. If allowedOrigins is empty, no CORS header is set (same-origin only). Otherwise, the request Origin is checked against the allowed list.

func RegisterReleaseRoutes added in v0.39.0 

func RegisterReleaseRoutes(mux *http.ServeMux, directory string)

RegisterReleaseRoutes uses only the configured local results root; the dashboard never supplies an arbitrary filesystem path or assesses summaries as ordinary runs. Each response re-verifies the collection artifacts.

func RegisterRoutes 

func RegisterRoutes(mux *http.ServeMux, store RunStore)

RegisterRoutes registers all web API routes on the given mux.

func RegisterRoutesWithStorage 

func RegisterRoutesWithStorage(mux *http.ServeMux, store RunStore, cfg *StorageConfig)

RegisterRoutesWithStorage registers all web API routes with storage config.

Types

type ConfidenceIntervalResponse 

type ConfidenceIntervalResponse struct {
	Lower           float64 `json:"lower"`
	Upper           float64 `json:"upper"`
	Mean            float64 `json:"mean"`
	ConfidenceLevel float64 `json:"confidenceLevel"`
}

ConfidenceIntervalResponse represents either a score or a binomial interval.

type ErrorResponse 

type ErrorResponse struct {
	Error string `json:"error"`
	Code  int    `json:"code"`
}

ErrorResponse is returned for errors.

type EvidenceRunResponse added in v0.39.0 

type EvidenceRunResponse struct {
	RunNumber    int                             `json:"runNumber"`
	Attempts     int                             `json:"attempts"`
	Cached       bool                            `json:"cached"`
	Assessment   string                          `json:"assessment"`
	Manifest     *models.EvidenceManifest        `json:"manifest,omitempty"`
	Explanations []models.RequirementExplanation `json:"explanations,omitempty"`
	Message      string                          `json:"message,omitempty"`
}

type FileStore 

type FileStore struct {
	// contains filtered or unexported fields
}

FileStore reads EvaluationOutcome JSON files from a directory.

func NewFileStore 

func NewFileStore(dir string) *FileStore

NewFileStore creates a FileStore that reads results from dir.

func (*FileStore) GetRun 

func (fs *FileStore) GetRun(id string) (*RunDetail, error)

GetRun returns a single run with full task details.

func (*FileStore) ListRuns 

func (fs *FileStore) ListRuns(sortField, order string) ([]RunSummary, error)

ListRuns returns all runs sorted by the given field and order.

func (*FileStore) Reload 

func (fs *FileStore) Reload() error

Reload forces a fresh reload of all result files from disk.

func (*FileStore) Summary 

func (fs *FileStore) Summary() (*SummaryResponse, error)

Summary returns aggregate metrics across all runs.

type GraderResult 

type GraderResult struct {
	Name    string  `json:"name"`
	Type    string  `json:"type"`
	Passed  bool    `json:"passed"`
	Score   float64 `json:"score"`
	Weight  float64 `json:"weight"`
	Message string  `json:"message"`
}

GraderResult is a single grader/validator result.

type Handlers 

type Handlers struct {
	// contains filtered or unexported fields
}

Handlers holds the HTTP handler methods for the web API.

func NewHandlers 

func NewHandlers(store RunStore) *Handlers

NewHandlers creates a new Handlers with the given store.

func NewHandlersWithStorage 

func NewHandlersWithStorage(store RunStore, cfg *StorageConfig) *Handlers

NewHandlersWithStorage creates a new Handlers with storage configuration.

func (*Handlers) HandleHealth 

func (h *Handlers) HandleHealth(w http.ResponseWriter, _ *http.Request)

HandleHealth returns a simple health check response.

func (*Handlers) HandleLabRuns added in v0.39.0 

func (h *Handlers) HandleLabRuns(w http.ResponseWriter, r *http.Request)

HandleLabRuns refreshes local artifacts for the opt-in dashboard.

func (*Handlers) HandleLatestRunEvents added in v0.38.0 

func (h *Handlers) HandleLatestRunEvents(w http.ResponseWriter, r *http.Request)

HandleLatestRunEvents preserves the legacy /api/events stream by replaying the newest run in the same SSE format as the v1 per-run endpoint.

func (*Handlers) HandleRunDetail 

func (h *Handlers) HandleRunDetail(w http.ResponseWriter, r *http.Request)

HandleRunDetail returns full run detail with per-task results.

func (*Handlers) HandleRunEvents added in v0.38.0 

func (h *Handlers) HandleRunEvents(w http.ResponseWriter, r *http.Request)

HandleRunEvents streams replayable Server-Sent Events for a run.

func (*Handlers) HandleRuns 

func (h *Handlers) HandleRuns(w http.ResponseWriter, r *http.Request)

HandleRuns returns a list of all runs, with optional sort/order query params.

func (*Handlers) HandleStorageStatus 

func (h *Handlers) HandleStorageStatus(w http.ResponseWriter, _ *http.Request)

HandleStorageStatus returns the current storage configuration status.

func (*Handlers) HandleSummary 

func (h *Handlers) HandleSummary(w http.ResponseWriter, _ *http.Request)

HandleSummary returns aggregate KPI metrics across all runs.

type HealthResponse 

type HealthResponse struct {
	Status  string `json:"status"`
	Version string `json:"version"`
}

HealthResponse is the health check response.

type ModelUsageResponse added in v0.38.9 

type ModelUsageResponse struct {
	Model            string   `json:"model"`
	AICredits        *float64 `json:"aiCredits,omitempty"`
	InputTokens      int      `json:"inputTokens"`
	CacheReadTokens  int      `json:"cacheReadTokens"`
	CacheWriteTokens int      `json:"cacheWriteTokens"`
	OutputTokens     int      `json:"outputTokens"`
}

ModelUsageResponse is the per-model usage breakdown for a run. AICredits is omitted when the Copilot SDK did not report a final AI-credit total for the model.

type PromptJudgeUsageResponse added in v0.39.0 

type PromptJudgeUsageResponse = judgeevidence.UsageMetadata

type PromptJudgingRunResponse added in v0.39.0 

type PromptJudgingRunResponse = judgeevidence.RunMetadata

type ReleaseCollection added in v0.39.0 

type ReleaseCollection struct {
	Path     string          `json:"path"`
	Decision ReleaseDecision `json:"decision"`
	Error    string          `json:"error,omitempty"`
}

type ReleaseDecision added in v0.39.0 

type ReleaseDecision struct {
	releasepolicy.Decision
	Kind  string                               `json:"kind"`
	Usage map[releasepolicy.Arm][]ReleaseUsage `json:"usage"`
}

type ReleaseUsage added in v0.39.0 

type ReleaseUsage struct {
	releasepolicy.UsageAxis
	Value *string `json:"value,omitempty"`
}

ReleaseUsage transports exact numeric tokens as strings so JavaScript does not round large integer token counts before rendering them.

type ReliabilitySummaryResponse added in v0.39.0 

type ReliabilitySummaryResponse struct {
	SuccessRateCI *ConfidenceIntervalResponse `json:"successRateCI,omitempty"`
	PassHatK      []float64                   `json:"passHatK,omitempty"`
}

ReliabilitySummaryResponse carries recorded estimates, never reconstructed ones.

type ResponderInfoResponse added in v0.37.0 

type ResponderInfoResponse struct {
	FollowupsSent int    `json:"followupsSent"`
	Outcome       string `json:"outcome"`
	Reason        string `json:"reason,omitempty"`
}

ResponderInfoResponse is the API representation of a responder-driven run summary.

type RunDetail 

type RunDetail struct {
	RunSummary
	Tasks        []TaskResult                `json:"tasks"`
	Statistics   *ReliabilitySummaryResponse `json:"statistics,omitempty"`
	RegradeUsage *PromptJudgeUsageResponse   `json:"regradeUsage,omitempty"`
}

RunDetail is the API response for a single run with per-task results.

func (RunDetail) MarshalJSON added in v0.39.0 

func (d RunDetail) MarshalJSON() ([]byte, error)

type RunEvent added in v0.38.0 

type RunEvent struct {
	SchemaVersion string       `json:"schemaVersion"`
	Sequence      int64        `json:"sequence"`
	RunID         string       `json:"runId"`
	Type          RunEventType `json:"type"`
	Data          RunEventData `json:"data"`
	Timestamp     time.Time    `json:"timestamp"`
}

RunEvent is the JSON payload sent in each Server-Sent Event.

type RunEventData added in v0.38.0 

type RunEventData struct {
	Spec           string  `json:"spec,omitempty"`
	Model          string  `json:"model,omitempty"`
	TaskName       string  `json:"taskName,omitempty"`
	Outcome        string  `json:"outcome,omitempty"`
	Score          float64 `json:"score,omitempty"`
	Duration       float64 `json:"duration,omitempty"`
	GraderName     string  `json:"graderName,omitempty"`
	GraderType     string  `json:"graderType,omitempty"`
	Passed         *bool   `json:"passed,omitempty"`
	Message        string  `json:"message,omitempty"`
	TotalTasks     int     `json:"totalTasks,omitempty"`
	CompletedTasks int     `json:"completedTasks,omitempty"`
	PassCount      int     `json:"passCount,omitempty"`
	FailCount      int     `json:"failCount,omitempty"`
	Tokens         int     `json:"tokens,omitempty"`
	Cost           float64 `json:"cost,omitempty"`
}

RunEventData carries event-specific progress details.

type RunEventType added in v0.38.0 

type RunEventType string

RunEventType identifies a live progress event in a run SSE stream. 

const (
	RunEventStarted       RunEventType = "run_started"
	RunEventTaskStarted   RunEventType = "task_started"
	RunEventStepExecuted  RunEventType = "step_executed"
	RunEventTaskCompleted RunEventType = "task_completed"
	RunEventCompleted     RunEventType = "run_completed"
	RunEventFailed        RunEventType = "run_failed"
)

type RunStore 

type RunStore interface {
	// ListRuns returns all runs, sorted by the given field and order.
	ListRuns(sortField, order string) ([]RunSummary, error)
	// GetRun returns a single run with full task details.
	GetRun(id string) (*RunDetail, error)
	// Summary returns aggregate metrics across all runs.
	Summary() (*SummaryResponse, error)
}

RunStore provides access to evaluation run data.

type RunSummary 

type RunSummary struct {
	ID              string  `json:"id"`
	Spec            string  `json:"spec"`
	Model           string  `json:"model"`
	JudgeModel      string  `json:"judgeModel,omitempty"`
	Outcome         string  `json:"outcome"`
	PassCount       int     `json:"passCount"`
	TaskCount       int     `json:"taskCount"`
	Tokens          int     `json:"tokens"`
	PremiumRequests float64 `json:"premiumRequests"`
	// AICredits is the final AI-credit total reported by the Copilot SDK for the
	// Copilot usage this run initiated. It is omitted for legacy artifacts and
	// runtimes that do not report final AI-credit metrics; clients must render
	// that as an unavailable state rather than substituting an estimate.
	AICredits *float64 `json:"aiCredits,omitempty"`
	// ModelUsage is the per-model usage breakdown for the run, sorted by model ID.
	ModelUsage []ModelUsageResponse `json:"modelUsage,omitempty"`
	Cost       float64              `json:"cost"`
	// CostSource records how Cost was computed: "sdk" (reported by the Copilot
	// SDK), "table" (priced from the embedded model rate table), or "estimate"
	// (flat-rate fallback). Empty for legacy summaries that carry no token/cost
	// data.
	CostSource    string    `json:"costSource,omitempty"`
	Duration      float64   `json:"duration"`
	Timestamp     time.Time `json:"timestamp"`
	Source        string    `json:"source,omitempty"` // "local" or "azure-blob"
	Skill         string    `json:"skill,omitempty"`
	Repetitions   int       `json:"repetitions,omitempty"`
	WeightedScore *float64  `json:"weightedScore,omitempty"`
	// contains filtered or unexported fields
}

RunSummary is the API response for a single run in the list.

func (RunSummary) MarshalJSON added in v0.39.0 

func (s RunSummary) MarshalJSON() ([]byte, error)

type SessionDigestResponse 

type SessionDigestResponse struct {
	ToolPolicyMode    string                    `json:"toolPolicyMode,omitempty"`
	ToolPolicyDenials []models.ToolPolicyDenial `json:"toolPolicyDenials,omitempty"`
	TotalTurns        int                       `json:"totalTurns"`
	ToolCallCount     int                       `json:"toolCallCount"`
	TokensIn          int                       `json:"tokensIn"`
	TokensOut         int                       `json:"tokensOut"`
	TokensTotal       int                       `json:"tokensTotal"`
	ToolsUsed         []string                  `json:"toolsUsed"`
	Errors            []string                  `json:"errors"`
}

SessionDigestResponse is the API representation of a session digest.

type StorageAdapter 

type StorageAdapter struct {
	// contains filtered or unexported fields
}

StorageAdapter adapts storage.ResultStore to the webapi.RunStore interface. It provides a bridge between the storage layer and the web API layer.

func NewStorageAdapter 

func NewStorageAdapter(store storage.ResultStore, source string) *StorageAdapter

NewStorageAdapter creates a RunStore backed by the given storage.ResultStore.

func (*StorageAdapter) GetRun 

func (sa *StorageAdapter) GetRun(id string) (*RunDetail, error)

GetRun returns a single run with full task details.

func (*StorageAdapter) ListRuns 

func (sa *StorageAdapter) ListRuns(sortField, order string) ([]RunSummary, error)

ListRuns returns all runs, sorted by the given field and order.

func (*StorageAdapter) Reload added in v0.39.0 

func (sa *StorageAdapter) Reload() error

Reload refreshes backends with local snapshot support; remote behavior is unchanged.

func (*StorageAdapter) Summary 

func (sa *StorageAdapter) Summary() (*SummaryResponse, error)

Summary returns aggregate metrics across all runs.

type StorageConfig 

type StorageConfig struct {
	Configured bool
	Provider   string
	Account    string
}

StorageConfig holds storage configuration for the status endpoint.

type StorageStatusResponse 

type StorageStatusResponse struct {
	Configured bool   `json:"configured"`
	Provider   string `json:"provider,omitempty"`
	Account    string `json:"account,omitempty"`
}

StorageStatusResponse is the storage configuration status.

type SummaryResponse 

type SummaryResponse struct {
	TotalRuns          int     `json:"totalRuns"`
	TotalTasks         int     `json:"totalTasks"`
	PassRate           float64 `json:"passRate"`
	AvgTokens          float64 `json:"avgTokens"`
	AvgPremiumRequests float64 `json:"avgPremiumRequests"`
	// AvgAICredits is the mean final AI-credit total across the runs that
	// reported one. It is omitted when no run carries AI-credit metrics.
	AvgAICredits *float64 `json:"avgAICredits,omitempty"`
	AvgCost      float64  `json:"avgCost"`
	// CostSource records the source of AvgCost across runs: "sdk", "table",
	// "estimate", or "mixed" when different runs were priced from different
	// sources. Empty when there are no runs, or when every aggregated run lacks
	// cost data (e.g. legacy ResultSummary rows that don't carry token usage).
	CostSource  string  `json:"costSource,omitempty"`
	AvgDuration float64 `json:"avgDuration"`
}

SummaryResponse is the aggregate KPI response.

type TaskResult 

type TaskResult struct {
	ID                 string                      `json:"id,omitempty"`
	Name               string                      `json:"name"`
	Prompt             string                      `json:"prompt,omitempty"`
	Outcome            string                      `json:"outcome"`
	Score              float64                     `json:"score"`
	WeightedScore      *float64                    `json:"weightedScore,omitempty"`
	Duration           float64                     `json:"duration"`
	GraderResults      []GraderResult              `json:"graderResults"`
	DependencyModes    []models.DependencyMode     `json:"dependencyModes,omitempty"`
	CommandInvocations []models.CommandInvocation  `json:"commandInvocations,omitempty"`
	Transcript         []TranscriptEventResponse   `json:"transcript,omitempty"`
	SessionDigest      *SessionDigestResponse      `json:"sessionDigest,omitempty"`
	Responder          *ResponderInfoResponse      `json:"responder,omitempty"`
	BootstrapCI        *ConfidenceIntervalResponse `json:"bootstrapCI,omitempty"`
	PassRateCI         *ConfidenceIntervalResponse `json:"passRateCI,omitempty"`
	Trials             *TrialCountsResponse        `json:"trials,omitempty"`
	EvidenceRuns       []EvidenceRunResponse       `json:"evidenceRuns,omitempty"`
	PromptJudgingRuns  []PromptJudgingRunResponse  `json:"promptJudgingRuns,omitempty"`
}

TaskResult is a per-task result within a run.

type TranscriptEventResponse 

type TranscriptEventResponse struct {
	Type       string `json:"type"`
	Content    string `json:"content,omitempty"`
	Message    string `json:"message,omitempty"`
	ToolCallID string `json:"toolCallId,omitempty"`
	ToolName   string `json:"toolName,omitempty"`
	Arguments  any    `json:"arguments,omitempty"`
	ToolResult any    `json:"toolResult,omitempty"`
	Success    *bool  `json:"success,omitempty"`
}

TranscriptEventResponse is the API representation of a transcript event.

type TrialCountsResponse added in v0.39.0 

type TrialCountsResponse struct {
	Passed int `json:"passed"`
	Total  int `json:"total"`
}

Source Files

View all Source files

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.