Feedback

SDK reference for feedback operations

Overview

Internal feedback and bug reporting system for beta usage

• Category: internal
• Auth Required: No
• Supported Modes: standard, delegated, service

Operations

reportBug

Report a bug with detailed context and reproduction steps

Arguments:

• title (string, required): Brief bug description
• description (string, required): Detailed description of the bug
• severity (string, required): Bug severity: critical, high, medium, or low
• stepsToReproduce (array, optional): Steps to reproduce the bug
• expectedBehavior (string, optional): What should happen
• actualBehavior (string, optional): What actually happens
• errorDetails (object, optional): Error details: { message, stack, code }
• context (object, optional): Additional context: { conversationId, recentMessages, platform, appVersion }
• llmAnalysis (string, optional): LLM analysis of the issue

Returns:

AdapterOperationResult - Normalized result with id, type, title, severity, createdAt, source

Response Fields:

Example:

const result = await mirra.feedback.reportBug({
  title: "example",
  description: "example",
  severity: "example"
});

reportToolFailure

Auto-report tool or adapter failures for debugging

Arguments:

• adapterType (string, required): Adapter type (e.g., jupiter, crypto)
• operation (string, required): Operation that failed (e.g., swap, sendToken)
• errorMessage (string, required): Error message from the failure
• errorCode (string, optional): Error code if available
• errorStack (string, optional): Error stack trace
• args (object, optional): Sanitized arguments that caused the failure
• llmAnalysis (string, optional): LLM analysis of why it failed
• suggestedFix (string, optional): LLM suggested fix
• context (object, optional): Additional context: { conversationId, userId, timestamp }

Returns:

AdapterOperationResult - Normalized result with id, type, adapterType, operation, errorMessage, createdAt, source

Response Fields:

Example:

const result = await mirra.feedback.reportToolFailure({
  adapterType: "example",
  operation: "example",
  errorMessage: "example"
});

reportMissingCapability

Report when LLM cannot fulfill a user request

Arguments:

• userRequest (string, required): What the user asked for
• reason (string, required): Why it could not be fulfilled
• suggestedCapability (string, optional): What capability would enable this
• relatedAdapters (array, optional): Adapters that might be relevant
• context (object, optional): Additional context: { conversationId }

Returns:

AdapterOperationResult - Normalized result with id, type, userRequest, reason, createdAt, source

Response Fields:

Example:

const result = await mirra.feedback.reportMissingCapability({
  userRequest: "example",
  reason: "example"
});

submitFeedback

Submit general user feedback

Arguments:

• sentiment (string, required): Sentiment: positive, negative, or neutral
• feedback (string, required): Feedback content
• category (string, optional): Category: ux, performance, feature, or general
• context (object, optional): Additional context: { feature, screen }

Returns:

AdapterOperationResult - Normalized result with id, type, sentiment, category, createdAt, source

Response Fields:

Example:

const result = await mirra.feedback.submitFeedback({
  sentiment: "example",
  feedback: "example"
});

submitFeatureRequest

Submit a feature request

Arguments:

• title (string, required): Feature title
• description (string, required): Feature description
• useCase (string, optional): Why the user needs this feature
• priority (string, optional): Priority: high, medium, or low

Returns:

AdapterOperationResult - Normalized result with id, type, title, priority, createdAt, source

Response Fields:

Example:

const result = await mirra.feedback.submitFeatureRequest({
  title: "example",
  description: "example"
});