Add A2AS Certificate

a2as.yaml

@@ -0,0 +1,215 @@
+manifest:
+  version: "0.1.2"
+  schema: https://a2as.org/cert/schema
+  subject:
+    name: mzurs/vistaiq
+    source: https://github.com/mzurs/vistaiq
+    branch: main
+    commit: "b28e74a7"
+    scope: [backend/agent/feedback_agent.py, backend/agent/insight_and_recommendations.py, backend/agent/market_research_report.py,
+      backend/agent/numerical_summary.py, backend/agent/routers/actionable_insights_and_recommendations.py, backend/agent/routers/analyze_feedbacks_by_file.py,
+      backend/agent/routers/config.py, backend/agent/routers/market_research_report.py, backend/agent/routers/numerical_summary.py,
+      backend/agent/routers/sentiment_analysis.py, backend/agent/routers/trend_identification.py, backend/agent/sentiment_analysis.py,
+      backend/agent/trend_indentification.py]
+  issued:
+    by: A2AS.org
+    at: '2026-01-26T16:15:55Z'
+  signatures:
+    digest: sha256:_1HrCW2ZlydPnYxJSy310OiKEVha3gsVPSbzj2nO7Pc
+    key: ed25519:hPT3S5Y9fT1rVgNDq2nVTclLJ8isc1XgViX4R9RFotM
+    sig: ed25519:5YhjhjFrzuoQo3s00EgkCgGVtOJR0J8dlWLoBD8ZMQVen7CJ0D1LRSLaaGcLX9vweN690WoqOAMW7X-6FCH2Ag
+
+agents:
+  feedback_agent:
+    type: instance
+    models: [model]
+    params:
+      name: FeedBack Summarizer Agent
+      output_type: FeedbackAgentResponse
+      instructions: ['1. Role & Expertise:', You are a professional market research and data analytics consultant., 'Your
+          goal is to deliver actionable insights based on the provided data, focusing on trends, customer sentiments, opportunities,
+          and risks.', 'Always structure responses for clarity, using concise, business-friendly language.', '2. Data Handling:',
+        'If given structured data (tables, survey results), summarize key findings first, then elaborate with analysis.',
+        'If given unstructured data (reviews, social media posts), extract main sentiments, recurring themes, and emerging
+          trends.', 'If keywords or topics are provided, research and summarize relevant market trends and competitive landscapes.',
+        3. Output Style, 'Use clear headings, bullet points, and short paragraphs for easy reading.', End every report with
+          “Key Recommendations” containing 3–5 actionable business steps., 'If uncertain about a fact, state assumptions transparently
+          rather than guessing.', '4. Depth & Balance:', 'Provide both qualitative insights (themes, sentiments) and quantitative
+          indicators (percentages, counts, frequencies) when possible.', Highlight both opportunities and risks., '5. Formatting
+          for Business Use:', Always provide an executive summary at the start for busy decision-makers., 'Use comparisons
+          where helpful (e.g., "Compared to last quarter…" or "This is 15% above industry average").', 'Suggest data visualizations
+          (charts, heatmaps) if applicable.']
+  insights_and_recommendations_agent:
+    type: instance
+    models: [model]
+    params:
+      name: Insights And Recommendations Agent
+      output_type: InsightsAndRecommendationsAgentResponse
+      instructions: ['You are an SME consultant using AI for market research. Analyze the following feedback or market data
+          to generate 4-6 actionable insights. For each insight:', '- Insight: State the key finding.', '- Evidence: Quote
+          or reference from the input.', '- Priority: Rate as urgent, important, or low.', '- Recommendation: Provide a practical
+          step for an SME, including estimated cost/effort (low/medium/high).', '- Potential outcome: Describe the expected
+          benefit (e.g., increased sales by 10-20%).', 'Input data: [input]', 'Insights:']
+  market_research_report_agent:
+    type: instance
+    models: [model]
+    params:
+      name: Market Research Report Agent
+      output_type: MarketResearchReportAgentResponse
+      instructions: ['You are a market research expert for SMEs. Create a structured report based on the following feedback,
+          survey data, or keywords. Include:', '1. Executive Summary: 1-2 paragraphs overview.', '2. Sentiment Breakdown:
+          Positive, negative, neutral with examples and overall score (-1 to +1).', '3. Key Trends: 3-5 trends with evidence
+          and SME impact.', '4. Actionable Insights: 4-6 insights with recommendations and potential outcomes.', '5. Conclusion:
+          Final advice for the SME.', 'Ensure the report is professional, concise (under 500 words), and focused on real-world
+          business value.', 'Input data: [input]', 'Report:']
+  numerical_summary_agent:
+    type: instance
+    models: [model]
+    params:
+      name: Numerical Summary Agent
+      output_type: AgentOutputSchema(NumericalSummaryAgentResponse, strict_json_schema=True)
+      instructions: ['You are a data analyst specializing in SME market research. Analyze the following customer feedback
+          or market data to produce a numerical summary. Provide:', '1. Sentiment Distribution: Percentage of positive, negative,
+          and neutral sentiments (sum to 100%).', '2. Keyword Frequency: Count of top 3-5 keywords or themes (e.g., "price,"
+          "delivery") mentioned in the input.', '3. Priority Counts: Number of urgent, important, and low-priority issues
+          identified.', '4. Numerical Insights: 2-3 additional metrics (e.g., average sentiment score [-1 to +1], percentage
+          of feedback mentioning specific topics).', 'Output the results in a structured JSON format, suitable for visualization.
+          Ensure calculations are accurate and based on the input data.', 'Input data: [input]', 'Output:', '{', '"sentiment_distribution":
+          {"positive": <percentage>, "negative": <percentage>, "neutral": <percentage>},', '"keyword_frequency": {"<keyword1>":
+          <count>, "<keyword2>": <count>, ...},', '"priority_counts": {"urgent": <count>, "important": <count>, "low": <count>},',
+        '"numerical_insights": {"<metric1>": <value>, "<metric2>": <value>, ...}', '}']
+  sentiment_analysis_agent:
+    type: instance
+    models: [model]
+    params:
+      name: Sentiment Analysis Agent
+      output_type: SentimentAnalysisAgentResponse
+      instructions: ['You are a sentiment analysis expert for SME market research. Analyze the following customer feedback
+          or survey data. Identify the overall sentiment (positive, negative, neutral, or mixed) and provide a breakdown:',
+        '- Positive elements: List 2-4 key positive points with examples.', '- Negative elements: List 2-4 key negative points
+          with examples.', '- Neutral elements: List any neutral observations.', '- Sentiment score: Rate overall sentiment
+          on a scale of -1 (very negative) to +1 (very positive).', '- Recommendations: Suggest 1-2 actionable steps for an
+          SME to improve based on the sentiments.']
+  trend_identification_agent:
+    type: instance
+    models: [model]
+    params:
+      name: Trend Identification Agent
+      output_type: TrendIndentificationAgentResponse
+      instructions: ['You are a trend analyst for SME market research. Review the following keywords, feedback, or market
+          data to identify 3-5 key trends. For each trend:', '- Describe the trend briefly.', '- Provide evidence from the
+          input.', '- Estimate impact: High, medium, or low for an SME.', '- Suggest opportunities: 1-2 ways an SME can capitalize
+          on or address the trend.', 'Input data: [input]', 'Trends:']
+
+models:
+  model:
+    type: variable
+    agents: [insights_and_recommendations_agent, market_research_report_agent, sentiment_analysis_agent, trend_identification_agent,
+      feedback_agent, numerical_summary_agent]
+
+imports:
+  actionable_insights_and_recommendations: agent.routers.actionable_insights_and_recommendations
+  Agent: agents.Agent
+  AgentOutputSchema: agents.AgentOutputSchema
+  analyze_feedbacks_by_file: agent.routers.analyze_feedbacks_by_file
+  APIRouter: fastapi.APIRouter
+  AsyncOpenAI: agents.AsyncOpenAI
+  BaseModel: pydantic.BaseModel
+  CORSMiddleware: fastapi.middleware.cors.CORSMiddleware
+  FastAPI: fastapi.FastAPI
+  feedback_agent: agent.feedback_agent.feedback_agent
+  FeedbackAgentResponse: agent.feedback_agent.FeedbackAgentResponse
+  File: fastapi.File
+  HTTPException: fastapi.HTTPException
+  insights_and_recommendations_agent: agent.insight_and_recommendations.insights_and_recommendations_agent
+  InsightsAndRecommendationsAgentResponse: agent.insight_and_recommendations.InsightsAndRecommendationsAgentResponse
+  io: io
+  load_dotenv: dotenv.load_dotenv
+  market_research_report: agent.routers.market_research_report
+  market_research_report_agent: agent.market_research_report.market_research_report_agent
+  MarketResearchReportAgentResponse: agent.market_research_report.MarketResearchReportAgentResponse
+  model_config: routers.config.model_config
+  numerical_summary: agent.routers.numerical_summary
+  numerical_summary_agent: agent.numerical_summary.numerical_summary_agent
+  NumericalSummaryAgentResponse: agent.numerical_summary.NumericalSummaryAgentResponse
+  OpenAIChatCompletionsModel: agents.OpenAIChatCompletionsModel
+  os: os
+  pd: pandas
+  pdfplumber: pdfplumber
+  RunConfig: agents.run.RunConfig
+  Runner: agents.Runner
+  sentiment_analysis: agent.routers.sentiment_analysis
+  sentiment_analysis_agent: agent.sentiment_analysis.sentiment_analysis_agent
+  SentimentAnalysisAgentResponse: agent.sentiment_analysis.SentimentAnalysisAgentResponse
+  trend_identification: agent.routers.trend_identification
+  trend_identification_agent: agent.trend_indentification.trend_identification_agent
+  TrendIndentificationAgentResponse: agent.trend_indentification.TrendIndentificationAgentResponse
+  UploadFile: fastapi.UploadFile
+  uvicorn: uvicorn
+
+functions:
+  analyze_feedback:
+    type: async
+    module: backend.agent.routers.analyze_feedbacks_by_file
+    args: [request]
+  analyze_file:
+    type: async
+    module: backend.agent.routers.analyze_feedbacks_by_file
+    args: [file]
+  insights_and_recommendations:
+    type: async
+    module: backend.agent.routers.actionable_insights_and_recommendations
+    args: [file]
+  is_valid_hex_key:
+    type: sync
+    module: backend.agent.routers.config
+    args: [key]
+    params:
+      returns: bool
+  market_research_report:
+    type: async
+    module: backend.agent.routers.market_research_report
+    args: [file]
+  model_config:
+    type: sync
+    module: backend.agent.routers.config
+  numerical_summary:
+    type: async
+    module: backend.agent.routers.numerical_summary
+    args: [file]
+  root:
+    type: async
+    module: backend.main
+  sentiment_analysis:
+    type: async
+    module: backend.agent.routers.sentiment_analysis
+    args: [file]
+  trend_identification:
+    type: async
+    module: backend.agent.routers.trend_identification
+    args: [file]
+
+variables:
+  GEMINI_API_KEY:
+    type: env
+    params:
+      caller: [os.getenv]
+      path: [backend.agent.routers.config]
+
+files:
+  pdf_file:
+    type: variable
+    actions: [read]
+    params:
+      caller: [pdfplumber.open]
+
+networks:
+  generativelanguage.googleapis.com:
+    type: api
+    actions: [GET]
+    urls: [/v1beta/openai/]
+    protocols: [https]
+    ports: ["443"]
+    params:
+      caller: [model_config]
+      links: [base_url]