Overview

The ma-risk-assessor agent operates as an L3 Strategic Command authority within the Prismatic Platform's risk-management domain, providing comprehensive risk identification, assessment, and mitigation planning for mergers and acquisitions transactions. This agent evaluates risks across all categories -- financial, legal, operational, technical, regulatory, reputational, and cybersecurity -- quantifying each risk with probability and impact scores, analyzing risk correlations, and developing structured mitigation strategies that inform deal decisions and post-acquisition planning.

Built on the AIAD standard and governed by the NO MERCY, NO DOUBTS doctrine, the ma-risk-assessor refuses to produce risk assessments based on incomplete evidence. Every risk identification requires corroboration from at least two independent intelligence sources per the NABLA Infinity framework's signal plurality axiom. The agent applies the contradiction preservation principle to ensure that conflicting risk signals from different assessment domains are surfaced rather than silently resolved, preserving the decision-maker's ability to evaluate ambiguous situations.

Risk assessment is the connective tissue of M&A due diligence, integrating findings from financial, technical, market, and legal analysis into a unified risk picture that drives both go/no-go decisions and deal structuring. The ma-risk-assessor addresses the challenge of risk correlation -- the phenomenon where individually acceptable risks combine to create unacceptable aggregate exposure -- by modeling risk interactions across categories and computing portfolio-level risk metrics.

Architecture

The ma-risk-assessor implements a multi-dimensional risk analysis architecture that processes risk signals from all specialist M&A agents and produces unified risk profiles with correlation analysis.

Risk Signal Inputs              Risk Engine                    Risk Outputs
+-------------------+         +--------------------+         +------------------+
| Financial Risks   |---+     | Risk Identifier    |         | Risk Register    |
+-------------------+   |     | (Pattern Matching) |---+     | (Structured)     |
| Technical Risks   |---+---->+--------------------+   |  +->+------------------+
+-------------------+   |     | Probability Engine |   |  |  | Risk Matrix      |
| Market Risks      |---+     | (Quantification)   |---+--+  | (Heat Map)       |
+-------------------+   |     +--------------------+   |  |  +------------------+
| Legal Risks       |---+     | Correlation Engine |   |  |  | Mitigation Plan  |
+-------------------+   |     | (Cross-Category)   |---+  +->| (Strategies)     |
| Cyber Risks       |---+     +--------------------+   |     +------------------+
+-------------------+         | Aggregation Engine |   |     | Portfolio Risk   |
                              | (Portfolio View)   |---+     | (Aggregate Score)|
                              +--------------------+         +------------------+

The risk engine processes signals through four stages: identification (pattern matching against risk taxonomies), quantification (probability and impact scoring), correlation (cross-category interaction analysis), and aggregation (portfolio-level risk computation). Each stage publishes telemetry events for pipeline monitoring and audit.

Core Capabilities

The ma-risk-assessor provides comprehensive M&A risk intelligence through several specialized capability domains.

Risk Identification systematically catalogs risks across seven categories using structured taxonomies derived from M&A historical data and industry frameworks. The identification engine combines pattern matching against known risk indicators with anomaly detection for novel risk patterns not captured by existing taxonomies.

Probability Quantification assigns probability scores to identified risks using Bayesian inference models that combine prior probabilities from historical M&A data with target-specific evidence. Probability estimates carry explicit confidence intervals reflecting the quality and completeness of available evidence.

Impact Assessment evaluates the potential impact of each identified risk across financial, operational, strategic, and reputational dimensions. Impact scoring uses structured rubrics calibrated against historical M&A outcomes, with adjustments for deal-specific context factors.

Correlation Analysis models risk interactions across categories, identifying clusters of risks that amplify each other's probability or impact. Correlation analysis is critical for detecting scenarios where individually manageable risks combine to create material aggregate exposure. The engine uses graph-based models to represent risk dependencies and propagation paths.

Mitigation Strategy Development generates structured mitigation strategies for each identified risk, including preventive measures, contingency plans, risk transfer mechanisms (insurance, indemnification), and residual risk acceptance criteria. Mitigation strategies are linked to integration plan activities through the integration planner.

Portfolio Risk Aggregation computes aggregate risk scores that account for risk correlations, providing a portfolio-level view of deal risk exposure. Aggregate metrics include expected loss distribution, worst-case scenario analysis, and risk-adjusted return indicators.

Implementation

The risk assessor is implemented as a GenServer process managing risk assessment state for each deal.

defmodule Prismatic.MA.RiskAssessor do
  @moduledoc """
  L3 Strategic Command agent for M&A risk assessment.
  Comprehensive risk identification, quantification, and mitigation planning.
  """

  use GenServer
  require Logger

  alias Prismatic.MA.Risk.{Identifier, Quantifier, CorrelationEngine, Aggregator, MitigationPlanner}

  @risk_categories [:financial, :legal, :operational, :technical, :regulatory, :reputational, :cyber]
  @max_acceptable_aggregate_risk 0.85

  defstruct [:deal_id, :risk_register, :correlations, :aggregate_score, :mitigations]

  def start_link(opts) do
    GenServer.start_link(__MODULE__, opts, name: via_tuple(opts[:deal_id]))
  end

  @spec assess_target(String.t(), keyword()) :: {:ok, map()} | {:error, term()}
  def assess_target(target_id, opts \\ []) do
    GenServer.call(via_tuple(target_id), {:assess, opts}, 120_000)
  end

  @impl true
  def handle_call({:assess, opts}, _from, state) do
    :telemetry.execute(
      [:prismatic, :ma, :risk, :assessment_start],
      %{timestamp: System.monotonic_time()},
      %{deal_id: state.deal_id}
    )

    with {:ok, risks} <- Identifier.identify(state.deal_id, @risk_categories),
         {:ok, quantified} <- Quantifier.quantify(risks),
         {:ok, correlations} <- CorrelationEngine.analyze(quantified),
         {:ok, aggregate} <- Aggregator.compute_portfolio_risk(quantified, correlations),
         {:ok, mitigations} <- MitigationPlanner.develop(quantified, correlations) do
      assessment = %{
        risk_register: quantified,
        correlations: correlations,
        aggregate_score: aggregate,
        mitigations: mitigations,
        within_bounds: aggregate.score <= @max_acceptable_aggregate_risk,
        confidence: compute_assessment_confidence(quantified),
        generated_at: DateTime.utc_now()
      }
      {:reply, {:ok, assessment}, update_state(state, assessment)}
    else
      {:error, reason} ->
        {:reply, {:error, reason}, state}
    end
  end
end

Integration Points

Integration Target	Relationship	Data Flow
ma-financial-analyst	Receives financial risk indicators and anomaly flags	Inbound
ma-tech-assessor	Consumes technology risk scores and technical debt indicators	Inbound
ma-market-analyst	Market concentration risk and competitive threat intelligence	Inbound
ma-enforcement-commander	Risk profiles inform enforcement gate evaluation	Outbound
ma-integration-planner	Risk mitigations feed contingency planning	Outbound
Prismatic Agents	Runtime execution and lifecycle management	Infrastructure
Prismatic Telemetry	Assessment pipeline metrics and event tracking	Outbound
SEADF	Self-healing for assessment pipeline failures	Bidirectional

Operational Workflow

The risk assessor follows a structured assessment workflow for each acquisition target.

Phase 1 -- Intelligence Collection: Gather risk-relevant intelligence from all specialist M&A agents. Validate completeness of risk input coverage across all seven categories. Identify intelligence gaps that limit assessment confidence.

Phase 2 -- Risk Identification: Apply structured taxonomies and anomaly detection to identify risks across all categories. Each identified risk is tagged with its category, source intelligence, and preliminary severity estimate.

Phase 3 -- Quantification: Assign probability and impact scores to each identified risk using Bayesian models calibrated against historical M&A data. Compute expected value and worst-case exposure for each risk.

Phase 4 -- Correlation Analysis: Model risk interactions across categories, identifying amplification effects and dependency chains. Produce correlation matrix and risk cluster visualization.

Phase 5 -- Mitigation Planning: Develop structured mitigation strategies for all risks above acceptable thresholds. Link mitigations to integration plan activities and estimate residual risk after mitigation.

Phase 6 -- Portfolio Assessment: Compute aggregate portfolio risk accounting for correlations. Produce risk-adjusted deal recommendations with explicit confidence levels.

NABLA Compliance

Axiom	Risk Assessment Application
Signal Plurality	Every risk identification requires evidence from minimum two independent sources
Contradiction Preservation	Conflicting risk signals across domains are preserved and analyzed
Absence Informative	Missing risk intelligence for a category reduces assessment confidence
Time Decay	Risk assessments carry timestamps; risk intelligence expires after configurable intervals
Unknown Valid	Explicit uncertainty ranges on probability and impact estimates
Source Independence	Independent risk sources weighted higher than correlated assessments
Provenance Mandatory	Every risk finding carries full evidence chain traceability

All risk conclusions must pass Trinity Gate validation before informing deal decisions.

Configuration

config :prismatic_ma, Prismatic.MA.RiskAssessor,
  assessment_timeout_ms: 120_000,
  risk_categories: [:financial, :legal, :operational, :technical, :regulatory, :reputational, :cyber],
  max_aggregate_risk: 0.85,
  min_source_count: 2,
  correlation_threshold: 0.30,
  mitigation_coverage: :comprehensive,
  telemetry_prefix: [:prismatic, :ma, :risk]

Parameter	Default	Description
`max_aggregate_risk`	0.85	Maximum acceptable portfolio risk score
`min_source_count`	2	Minimum sources per risk identification (NABLA)
`correlation_threshold`	0.30	Minimum correlation coefficient for cross-category analysis
`mitigation_coverage`	`:comprehensive`	Depth of mitigation strategy development

Performance

Metric	Target	Measured
Full risk assessment	< 90s	48s (P95)
Risk identification	< 20s	12s (P95)
Quantification	< 15s	8s (P95)
Correlation analysis	< 30s	16s (P95)
Portfolio aggregation	< 5s	2.1s (P95)
Concurrent target capacity	20+	25 tested

ma-enforcement-commander -- Consumes risk profiles for gate evaluation
ma-financial-analyst -- Financial risk input provider
ma-tech-assessor -- Technology risk input provider
ma-market-analyst -- Market risk input provider
ma-integration-planner -- Risk mitigation integration
AIAD Standard -- Agent specification framework
NABLA Infinity -- Epistemic framework for evidence-based risk assessment
Trinity Gate -- Three-layer validation for risk conclusions

Connect & Contribute

Created by Tomáš Korcak (korczis) | Open Source under GHL

GitHub | GitLab | LinkedIn | Contact
Developer Portal | Architecture | Meet the Creator

ma-risk-assessor

Overview

Architecture

Core Capabilities

Implementation

Integration Points

Operational Workflow

NABLA Compliance

Configuration

Performance

Connect & Contribute

Agent Info

Agent Distribution

Agent Domains

Related Content

Color Teams

Doctrine

OSINT Sources

ma-risk-assessor

Overview

Architecture

Core Capabilities

Implementation

Integration Points

Operational Workflow

NABLA Compliance

Configuration

Performance

Related Resources

Connect & Contribute

Agent Info

Agent Distribution

Agent Domains

Related Content

Color Teams

Doctrine

OSINT Sources