Security & Trust Policies¶

Every tool invocation in SynthOrg passes through the SecOps security pipeline. This guide covers how to configure autonomy levels, trust strategies, approval workflows, custom policies, and output scanning. For the internal architecture of the security subsystem, see the Security reference.

Autonomy Levels¶

Autonomy levels control which actions require human approval. Set the company-wide level in config.autonomy.level, with optional per-agent overrides:

config:
  autonomy:
    level: semi

agents:
  - role: "Junior Developer"
    autonomy_level: supervised  # more restrictive than company default
  - role: "CEO"
    autonomy_level: full        # less restrictive than company default

Trust Strategies¶

Trust strategies control how agents earn (or lose) access to higher-privilege tool categories over time. Configure via the trust section:

trust:
  strategy: disabled
  initial_level: standard

trust:
  strategy: weighted
  initial_level: restricted
  weights:
    task_difficulty: 0.3
    completion_rate: 0.25
    error_rate: 0.25
    human_feedback: 0.2
  promotion_thresholds:
    restricted_to_standard:
      score: 0.7
      requires_human_approval: false
    standard_to_elevated:
      score: 0.9
      requires_human_approval: true  # REQUIRED (security invariant)

trust:
  strategy: per_category
  initial_level: restricted
  initial_category_levels:
    code: restricted
    vcs: sandboxed
    deploy: sandboxed
  category_criteria:
    code:
      restricted_to_standard:
        tasks_completed: 10
        quality_score_min: 7.0
        requires_human_approval: false
      standard_to_elevated:
        tasks_completed: 50
        quality_score_min: 8.5
        requires_human_approval: true  # REQUIRED (security invariant)
    vcs:
      sandboxed_to_restricted:
        tasks_completed: 5
        quality_score_min: 7.0

trust:
  strategy: milestone
  initial_level: restricted
  milestones:
    restricted_to_standard:
      tasks_completed: 20
      quality_score_min: 7.5
      time_active_days: 7
      clean_history_days: 3
      auto_promote: true
      requires_human_approval: false
    standard_to_elevated:
      tasks_completed: 100
      quality_score_min: 8.5
      time_active_days: 30
      clean_history_days: 14
      auto_promote: false
      requires_human_approval: true  # REQUIRED (security invariant)
  re_verification:
    enabled: true
    interval_days: 90
    decay_on_idle_days: 30
    decay_on_error_rate: 0.15

Tool Access Levels¶

Trust levels map to tool access categories:

Levels form a hierarchy where each includes all categories from lower levels.

Re-verification¶

For the milestone strategy, re-verification periodically re-evaluates trust:

Security Configuration¶

The security section controls the SecOps rule engine, output scanning, and audit logging:

security:
  enabled: true
  audit_enabled: true
  post_tool_scanning_enabled: true
  output_scan_policy_type: autonomy_tiered
  hard_deny_action_types:
    - "deploy:production"
    - "db:admin"
    - "org:fire"
  auto_approve_action_types:
    - "code:read"
    - "docs:write"

Security Fields¶

Rule Engine¶

The rule engine runs synchronous checks against every tool invocation:

security:
  rule_engine:
    credential_patterns_enabled: true
    data_leak_detection_enabled: true
    destructive_op_detection_enabled: true
    path_traversal_detection_enabled: true
    max_argument_length: 100000
    custom_allow_bypasses_detectors: false

Built-in Detectors¶

Each detector can be independently enabled or disabled.

Custom Security Policies¶

Define custom rules to allow, deny, or escalate specific action types:

security:
  custom_policies:
    - name: "block-external-comms"
      description: "Prevent agents from sending external communications"
      action_types:
        - "comms:external"
      verdict: deny
      risk_level: high
      enabled: true
    - name: "escalate-deploys"
      description: "Escalate staging deployments for review"
      action_types:
        - "deploy:staging"
      verdict: escalate
      risk_level: medium

Policy Rule Fields¶

Action Types¶

Action types follow a category:action format. Built-in types include:

LLM Security Fallback¶

For actions that the rule engine cannot classify with high confidence, an LLM from a different provider family can provide cross-validation:

security:
  llm_fallback:
    enabled: true
    model: "example-medium-001"
    timeout_seconds: 10.0
    max_input_tokens: 2000
    on_error: escalate
    reason_visibility: generic
    argument_truncation: per_value

Output Scanning¶

After tool execution, the output scanner checks for leaked secrets and PII:

security:
  output_scan_policy_type: autonomy_tiered

See Also¶

• Company Configuration -- full configuration reference
• Security -- security architecture reference
• Design: Operations -- security design specification