@panguard-ai/atr 1.3.0 → 1.4.1

package/rules/excessive-autonomy/{ATR-2026-050-runaway-agent-loop.yaml → ATR-2026-00050-runaway-agent-loop.yaml}

@@ -1,5 +1,6 @@
-title: 'Runaway Agent Loop Detection'
-id: ATR-2026-050
+title: "Runaway Agent Loop Detection"
+id: ATR-2026-00050
+rule_version: 1
 status: experimental
 description: |
   Detects when an agent enters a runaway loop through repeated identical
@@ -9,26 +10,27 @@ description: |
   descriptions, recursive invocation patterns, and stalled progress
   indicators. Runaway loops waste computational resources, accumulate
   costs, and may indicate logic errors or adversarial manipulation.
-author: 'ATR Community'
-date: '2026/03/08'
-schema_version: '0.1'
+author: "ATR Community"
+date: "2026/03/08"
+schema_version: "0.1"
 detection_tier: pattern
 maturity: experimental
 severity: high
 
 references:
   owasp_llm:
-    - 'LLM06:2025 - Excessive Agency'
-    - 'LLM10:2025 - Unbounded Consumption'
+    - "LLM06:2025 - Excessive Agency"
+    - "LLM10:2025 - Unbounded Consumption"
   owasp_agentic:
-    - 'ASI05:2026 - Unexpected Code Execution'
+    - "ASI05:2026 - Unexpected Code Execution"
   mitre_atlas:
-    - 'AML.T0053 - LLM Plugin Compromise'
-    - 'AML.T0046 - Spamming ML System with Chaff Data'
+    - "AML.T0053 - LLM Plugin Compromise"
+    - "AML.T0046 - Spamming ML System with Chaff Data"
 
 tags:
   category: excessive-autonomy
   subcategory: runaway-loop
+  scan_target: mcp
   confidence: medium
 
 agent_source:
@@ -43,57 +45,57 @@ detection:
     - field: content
       operator: regex
       value: "(?i)(?:retrying|retry\\s+(?:attempt|number|#|count))\\s*(?::?\\s*)?\\d+"
-      description: 'Retry counter pattern indicating repeated attempts'
+      description: "Retry counter pattern indicating repeated attempts"
     - field: content
       operator: regex
       value: "(?i)attempt\\s+\\d+\\s+(?:of|/)\\s+\\d+"
-      description: 'Attempt X of Y counter pattern (e.g., attempt 5 of 10)'
+      description: "Attempt X of Y counter pattern (e.g., attempt 5 of 10)"
     - field: content
       operator: regex
       value: "(?i)(?:trying\\s+again|let\\s+me\\s+(?:try|attempt)\\s+(?:again|once\\s+more|one\\s+more\\s+time)|retrying\\s+(?:the\\s+)?(?:same|previous|last)\\s+(?:operation|request|call|action))"
-      description: 'Natural language retry patterns in agent output'
+      description: "Natural language retry patterns in agent output"
     - field: content
       operator: regex
       value: "(?i)(?:still\\s+(?:failing|not\\s+working|getting\\s+(?:the\\s+same\\s+)?error)|same\\s+error\\s+(?:again|as\\s+before|persists)|error\\s+(?:persists|continues|remains))"
-      description: 'Stalled progress with persistent error indicators'
+      description: "Stalled progress with persistent error indicators"
     - field: content
       operator: regex
       value: "(?i)(?:calling\\s+(?:myself|this\\s+(?:function|tool|agent))\\s+(?:again|recursively)|recursive\\s+(?:call|invocation|execution)|self[_-]?(?:invoke|call|trigger))"
-      description: 'Recursive self-invocation patterns'
+      description: "Recursive self-invocation patterns"
     - field: content
       operator: regex
       value: "(?i)(?:iteration\\s+(?:#\\s*)?\\d{2,}|loop\\s+(?:iteration|count|index)\\s*[:=]?\\s*\\d{2,}|cycle\\s+\\d{2,})"
-      description: 'High iteration count indicating prolonged looping'
+      description: "High iteration count indicating prolonged looping"
     - field: content
       operator: regex
       value: "(?i)(?:repeating\\s+(?:the\\s+)?(?:same|previous|identical)\\s+(?:step|action|operation|request)|performing\\s+(?:the\\s+)?(?:same|identical)\\s+(?:action|call)\\s+again)"
-      description: 'Agent acknowledging it is repeating identical actions'
+      description: "Agent acknowledging it is repeating identical actions"
     - field: content
       operator: regex
       value: "(?i)(?:max(?:imum)?\\s+retries?\\s+(?:reached|exceeded|hit)|(?:exceeded|hit|reached)\\s+(?:the\\s+)?(?:retry|attempt|iteration)\\s+(?:limit|maximum|cap))"
-      description: 'Retry limit reached indicators'
+      description: "Retry limit reached indicators"
     - field: content
       operator: regex
       value: "(?i)(?:no\\s+progress\\s+(?:after|in)\\s+\\d+\\s+(?:attempts?|tries|iterations?|cycles?)|stuck\\s+(?:in\\s+(?:a\\s+)?loop|on\\s+(?:the\\s+)?same\\s+(?:step|error)))"
-      description: 'Explicit stall or no-progress acknowledgement'
+      description: "Explicit stall or no-progress acknowledgement"
     - field: content
       operator: regex
       value: "(?i)(?:(?:while|for)\\s*\\(\\s*(?:true|1|;;)\\s*\\)|(?:loop|repeat)\\s*\\{\\s*(?:call|invoke|execute))"
-      description: 'Infinite loop constructs in generated or executed code'
+      description: "Infinite loop constructs in generated or executed code"
     - field: content
       operator: regex
       value: "(?i)(?:will\\s+keep\\s+(?:trying|retrying|attempting)|(?:continuously|endlessly|infinitely)\\s+(?:retrying|looping|repeating|calling))"
-      description: 'Agent declaring intent to retry indefinitely'
+      description: "Agent declaring intent to retry indefinitely"
     - field: content
       operator: regex
       value: "(?i)(?:spawn(?:ing|ed)?\\s+(?:another|new|additional)\\s+(?:instance|copy|clone)\\s+of\\s+(?:myself|this\\s+agent)|fork(?:ing|ed)?\\s+(?:a\\s+)?(?:new\\s+)?(?:agent|process|instance))"
-      description: 'Agent spawning copies of itself (fork bomb pattern)'
+      description: "Agent spawning copies of itself (fork bomb pattern)"
   condition: any
   false_positives:
-    - 'Legitimate batch processing with progress counters in output'
-    - 'Polling agents designed to check status at regular intervals'
-    - 'Long-running data pipeline agents with expected high iteration counts'
-    - 'Retry logic in error handling with bounded retry limits'
+    - "Legitimate batch processing with progress counters in output"
+    - "Polling agents designed to check status at regular intervals"
+    - "Long-running data pipeline agents with expected high iteration counts"
+    - "Retry logic in error handling with bounded retry limits"
 
 response:
   actions:
@@ -110,25 +112,25 @@ response:
 
 test_cases:
   true_positives:
-    - input: 'Retrying attempt 15 of 10. The request still returns the same error.'
+    - input: "Retrying attempt 15 of 10. The request still returns the same error."
       expected: triggered
-      description: 'Retry counter exceeding stated maximum'
-    - input: 'Still failing after 25 iterations. Let me try again with the same parameters.'
+      description: "Retry counter exceeding stated maximum"
+    - input: "Still failing after 25 iterations. Let me try again with the same parameters."
       expected: triggered
-      description: 'Agent stuck in retry loop with no progress'
-    - input: 'Calling myself recursively to process the remaining items. Recursive call depth: 47.'
+      description: "Agent stuck in retry loop with no progress"
+    - input: "Calling myself recursively to process the remaining items. Recursive call depth: 47."
       expected: triggered
-      description: 'Recursive self-invocation with deep call stack'
-    - input: 'Will keep trying until it works. Continuously retrying the API call.'
+      description: "Recursive self-invocation with deep call stack"
+    - input: "Will keep trying until it works. Continuously retrying the API call."
       expected: triggered
-      description: 'Agent declaring intent to retry indefinitely'
+      description: "Agent declaring intent to retry indefinitely"
   true_negatives:
-    - input: 'Processing batch 3 of 5. All items completed successfully so far.'
+    - input: "Processing batch 3 of 5. All items completed successfully so far."
       expected: not_triggered
-      description: 'Legitimate batch processing with bounded progress counter'
-    - input: 'The API returned an error. Retrying once with exponential backoff.'
+      description: "Legitimate batch processing with bounded progress counter"
+    - input: "The API returned an error. Retrying once with exponential backoff."
       expected: not_triggered
-      description: 'Normal bounded retry logic in error handling'
-    - input: 'Task completed after 3 attempts. Here are the final results.'
+      description: "Normal bounded retry logic in error handling"
+    - input: "Task completed after 3 attempts. Here are the final results."
       expected: not_triggered
-      description: 'Successful completion after reasonable retry count'
+      description: "Successful completion after reasonable retry count"

package/rules/excessive-autonomy/{ATR-2026-051-resource-exhaustion.yaml → ATR-2026-00051-resource-exhaustion.yaml}

@@ -1,5 +1,6 @@
-title: 'Agent Resource Exhaustion Detection'
-id: ATR-2026-051
+title: "Agent Resource Exhaustion Detection"
+id: ATR-2026-00051
+rule_version: 1
 status: experimental
 description: |
   Detects when an agent causes resource exhaustion through bulk operations,
@@ -9,26 +10,27 @@ description: |
   as SELECT * without LIMIT, mass iteration directives, unbounded batch
   sizes, and fork/spawn patterns that can degrade system performance or
   cause denial of service.
-author: 'ATR Community'
-date: '2026/03/08'
-schema_version: '0.1'
+author: "ATR Community"
+date: "2026/03/08"
+schema_version: "0.1"
 detection_tier: pattern
 maturity: experimental
 severity: high
 
 references:
   owasp_llm:
-    - 'LLM06:2025 - Excessive Agency'
-    - 'LLM10:2025 - Unbounded Consumption'
+    - "LLM06:2025 - Excessive Agency"
+    - "LLM10:2025 - Unbounded Consumption"
   owasp_agentic:
-    - 'ASI05:2026 - Unexpected Code Execution'
+    - "ASI05:2026 - Unexpected Code Execution"
   mitre_atlas:
-    - 'AML.T0046 - Spamming ML System with Chaff Data'
-    - 'AML.T0053 - LLM Plugin Compromise'
+    - "AML.T0046 - Spamming ML System with Chaff Data"
+    - "AML.T0053 - LLM Plugin Compromise"
 
 tags:
   category: excessive-autonomy
   subcategory: resource-exhaustion
+  scan_target: mcp
   confidence: medium
 
 agent_source:
@@ -43,57 +45,57 @@ detection:
     - field: content
       operator: regex
       value: "(?i)SELECT\\s+\\*\\s+FROM\\s+\\w+(?:\\s*;|\\s*$|\\s+WHERE)(?!.*\\bLIMIT\\b)"
-      description: 'Unbounded SELECT * query without LIMIT clause'
+      description: "Unbounded SELECT * query without LIMIT clause"
     - field: content
       operator: regex
       value: "(?i)(?:for\\s+each|iterate\\s+(?:over\\s+)?(?:all|every)|process\\s+(?:all|every|each)\\s+(?:record|row|entry|item|file|document))"
-      description: 'Bulk iteration directive over entire dataset'
+      description: "Bulk iteration directive over entire dataset"
     - field: content
       operator: regex
       value: "(?i)(?:limit|max_results?|page_size|batch_size|count|top)\\s*[:=]\\s*(?:\\d{5,}|all|unlimited|-1|999+|MAX_INT|infinity)"
-      description: 'Excessively large or unbounded query parameters'
+      description: "Excessively large or unbounded query parameters"
     - field: content
       operator: regex
       value: "(?i)(?:delete\\s+(?:all|every|each)\\s+(?:files?|records?|entries|rows?|documents?)|rm\\s+-rf\\s+[/*]|truncate\\s+(?:all\\s+)?(?:tables?|data))"
-      description: 'Mass deletion or truncation operations'
+      description: "Mass deletion or truncation operations"
     - field: content
       operator: regex
       value: "(?i)(?:download(?:ing)?\\s+(?:all|every|the\\s+entire)|(?:fetch(?:ing)?|pull(?:ing)?|retriev(?:e|ing))\\s+(?:all|every|the\\s+complete|the\\s+entire)\\s+(?:dataset|database|table|collection|bucket))"
-      description: 'Bulk data download or retrieval of entire datasets'
+      description: "Bulk data download or retrieval of entire datasets"
     - field: content
       operator: regex
       value: "(?i)(?:send\\s+(?:a\\s+)?(?:request|message|email|notification)\\s+to\\s+(?:all|every|each)\\s+(?:user|member|contact|subscriber|endpoint))"
-      description: 'Mass messaging or notification blast patterns'
+      description: "Mass messaging or notification blast patterns"
     - field: content
       operator: regex
       value: "(?i)(?:spawn(?:ing)?\\s+\\d{2,}\\s+(?:(?:concurrent|parallel|simultaneous)\\s+)?(?:threads?|processes?|workers?|instances?|tasks?|connections?)|(?:create|open|launch)\\s+\\d{2,}\\s+(?:concurrent|parallel|simultaneous))"
-      description: 'Spawning excessive concurrent processes or connections'
+      description: "Spawning excessive concurrent processes or connections"
     - field: content
       operator: regex
       value: "(?i)(?:while\\s*\\(\\s*true\\s*\\)|for\\s*\\(\\s*;\\s*;\\s*\\)|do\\s*\\{).*(?:fetch|request|call|query|send|write)"
-      description: 'Infinite loop with resource-consuming operations inside'
+      description: "Infinite loop with resource-consuming operations inside"
     - field: content
       operator: regex
       value: "(?i)(?:(?:scrape|crawl|spider|harvest)\\s+(?:all|every|the\\s+entire)\\s+(?:site|website|domain|pages?)|recursive(?:ly)?\\s+(?:crawl|scrape|fetch|download))"
-      description: 'Unbounded web scraping or crawling operations'
+      description: "Unbounded web scraping or crawling operations"
     - field: content
       operator: regex
       value: "(?i)(?:no\\s+(?:limit|cap|maximum|bound|restriction)\\s+(?:on|for)\\s+(?:requests?|queries|calls?|operations?|iterations?)|(?:unlimited|unbounded|uncapped)\\s+(?:requests?|queries|calls?|operations?))"
-      description: 'Explicit removal of rate limits or resource caps'
+      description: "Explicit removal of rate limits or resource caps"
     - field: content
       operator: regex
       value: "(?i)(?:process(?:ing)?\\s+(?:the\\s+)?(?:entire|full|complete|whole)\\s+(?:corpus|dataset|collection|index|archive)|scan(?:ning)?\\s+(?:all|every)\\s+\\d{4,}\\s+(?:records?|files?|entries))"
-      description: 'Full corpus processing or scanning massive datasets'
+      description: "Full corpus processing or scanning massive datasets"
     - field: content
       operator: regex
       value: "(?i)(?:copy\\s+(?:all|every|entire)\\s+(?:data|files?|records?|tables?)\\s+(?:to|into)|replicate\\s+(?:the\\s+)?(?:entire|full|complete)\\s+(?:database|dataset|collection))"
-      description: 'Mass data replication or copy operations'
+      description: "Mass data replication or copy operations"
   condition: any
   false_positives:
-    - 'Authorized large-scale data processing or ETL pipeline agents'
-    - 'Batch analysis tasks with expected high volume approved by admin'
-    - 'Initial data ingestion or indexing operations with known high volume'
-    - 'Database export operations authorized by data team'
+    - "Authorized large-scale data processing or ETL pipeline agents"
+    - "Batch analysis tasks with expected high volume approved by admin"
+    - "Initial data ingestion or indexing operations with known high volume"
+    - "Database export operations authorized by data team"
 
 response:
   actions:
@@ -110,28 +112,28 @@ response:
 
 test_cases:
   true_positives:
-    - input: 'SELECT * FROM transactions;'
+    - input: "SELECT * FROM transactions;"
       expected: triggered
-      description: 'Unbounded SELECT query without LIMIT'
-    - input: 'I will iterate over all 500000 records and process each one individually.'
+      description: "Unbounded SELECT query without LIMIT"
+    - input: "I will iterate over all 500000 records and process each one individually."
       expected: triggered
-      description: 'Bulk iteration over massive dataset'
-    - input: 'Setting batch_size=999999 and limit=unlimited for maximum throughput.'
+      description: "Bulk iteration over massive dataset"
+    - input: "Setting batch_size=999999 and limit=unlimited for maximum throughput."
       expected: triggered
-      description: 'Excessively large batch parameters'
-    - input: 'Downloading all files from the S3 bucket. Fetching the entire dataset for local processing.'
+      description: "Excessively large batch parameters"
+    - input: "Downloading all files from the S3 bucket. Fetching the entire dataset for local processing."
       expected: triggered
-      description: 'Bulk data download of entire storage'
-    - input: 'Spawning 50 concurrent workers to send requests to the API.'
+      description: "Bulk data download of entire storage"
+    - input: "Spawning 50 concurrent workers to send requests to the API."
       expected: triggered
-      description: 'Excessive concurrent process spawning'
+      description: "Excessive concurrent process spawning"
   true_negatives:
-    - input: 'SELECT id, name FROM users WHERE active=true LIMIT 100;'
+    - input: "SELECT id, name FROM users WHERE active=true LIMIT 100;"
       expected: not_triggered
-      description: 'Bounded query with column selection and LIMIT'
-    - input: 'Processing the 25 records from the filtered result set.'
+      description: "Bounded query with column selection and LIMIT"
+    - input: "Processing the 25 records from the filtered result set."
       expected: not_triggered
-      description: 'Processing a reasonably sized result set'
-    - input: 'Fetching the first page of results with page_size=20.'
+      description: "Processing a reasonably sized result set"
+    - input: "Fetching the first page of results with page_size=20."
       expected: not_triggered
-      description: 'Paginated query with reasonable page size'
+      description: "Paginated query with reasonable page size"

package/rules/excessive-autonomy/{ATR-2026-052-cascading-failure.yaml → ATR-2026-00052-cascading-failure.yaml}

@@ -1,5 +1,6 @@
-title: 'Cascading Failure Detection in Agent Pipelines'
-id: ATR-2026-052
+title: "Cascading Failure Detection in Agent Pipelines"
+id: ATR-2026-00052
+rule_version: 1
 status: experimental
 description: |
   Detects cascading failure patterns in automated agent pipelines where
@@ -10,26 +11,27 @@ description: |
   incorrect signals. These patterns exploit the "trust the previous
   stage" assumption in multi-step agent workflows.
   Note: This rule detects textual descriptions of cascading failure patterns, not live cascading failures. Structural cascade prevention requires behavioral monitoring.
-author: 'ATR Community'
-date: '2026/03/09'
-schema_version: '0.1'
+author: "ATR Community"
+date: "2026/03/09"
+schema_version: "0.1"
 detection_tier: pattern
 maturity: experimental
 severity: high
 
 references:
   owasp_llm:
-    - 'LLM06:2025 - Excessive Agency'
-    - 'LLM05:2025 - Improper Output Handling'
+    - "LLM06:2025 - Excessive Agency"
+    - "LLM05:2025 - Improper Output Handling"
   owasp_agentic:
-    - 'ASI08:2026 - Cascading Failures'
+    - "ASI08:2026 - Cascading Failures"
   mitre_atlas:
-    - 'AML.T0053 - LLM Plugin Compromise'
-    - 'AML.T0046 - Spamming ML System with Chaff Data'
+    - "AML.T0053 - LLM Plugin Compromise"
+    - "AML.T0046 - Spamming ML System with Chaff Data"
 
 tags:
   category: excessive-autonomy
   subcategory: cascading-failure
+  scan_target: mcp
   confidence: medium
 
 agent_source:
@@ -44,53 +46,53 @@ detection:
     - field: content
       operator: regex
       value: "(?i)auto[_-]?approv(?:ed|ing|al).{0,30}(?:without|no|skipping)\\s+(?:human|manual|review|check|verification|validation)"
-      description: 'Automated approval bypassing human review checkpoint'
+      description: "Automated approval bypassing human review checkpoint"
     - field: content
       operator: regex
       value: "(?i)(?:propagat(?:ed|ing)|forward(?:ed|ing)|pass(?:ed|ing))\\s+(?:error|failure|exception|fault)\\s+(?:to|through|downstream|to\\s+next\\s+(?:stage|step|agent))"
-      description: 'Error propagating downstream through pipeline stages'
+      description: "Error propagating downstream through pipeline stages"
     - field: content
       operator: regex
       value: "(?i)(?:stage|step|phase)\\s+\\d+\\s+(?:failed|error|timed\\s+out).*(?:stage|step|phase)\\s+\\d+\\s+(?:also|subsequently|then)\\s+(?:failed|errored)"
-      description: 'Multiple pipeline stages failing in sequence'
+      description: "Multiple pipeline stages failing in sequence"
     - field: content
       operator: regex
       value: "(?i)(?:trusting|accepting|using)\\s+(?:previous|upstream|prior)\\s+(?:stage|step|agent)\\s+(?:output|result|response)\\s+(?:without|no)\\s+(?:validation|verification|check)"
-      description: 'Stage consuming upstream output without validation'
+      description: "Stage consuming upstream output without validation"
     - field: content
       operator: regex
       value: "(?i)(?:rollback|revert|undo|destroy|delete|terminate)\\s+(?:all|everything|entire|production|live).{0,40}(?:based\\s+on|triggered\\s+by|due\\s+to)\\s+(?:automated|agent|AI|pipeline)"
-      description: 'Destructive action triggered by automated pipeline signal'
+      description: "Destructive action triggered by automated pipeline signal"
     - field: content
       operator: regex
       value: "(?i)(?:amplif(?:ied|ying)|escalat(?:ed|ing)|magnif(?:ied|ying))\\s+(?:the\\s+)?(?:error|issue|problem|signal|impact|damage)\\s+(?:across|through|to)\\s+(?:multiple|all|downstream)"
-      description: 'Error amplification across multiple pipeline stages'
+      description: "Error amplification across multiple pipeline stages"
     - field: content
       operator: regex
       value: "(?i)(?:no\\s+(?:human|manual)\\s+(?:in|checkpoint|review|gate|approval)\\s+(?:the|this)?\\s*(?:loop|pipeline|chain|workflow)|(?:fully|completely)\\s+automated\\s+(?:pipeline|workflow|chain)\\s+(?:with|including)\\s+(?:destructive|production|critical))"
-      description: 'Fully automated pipeline with no human-in-the-loop for critical actions'
+      description: "Fully automated pipeline with no human-in-the-loop for critical actions"
     - field: content
       operator: regex
       value: "(?i)(?:blind(?:ly)?|unconditional(?:ly)?|automatic(?:ally)?)\\s+(?:executing|running|applying|deploying).{0,60}(?:previous|upstream|AI|agent|automated)(?:\\s+(?:AI|agent|system|model))?"
-      description: 'Blind execution of upstream agent recommendations'
+      description: "Blind execution of upstream agent recommendations"
     - field: content
       operator: regex
       value: "(?i)(?!(?:post[- ]?(?:incident|mortem)|review|analysis|report|last\\s+(?:week|month|time)|caused\\s+by|root\\s+cause|lessons?\\s+learned|investigation|retrospective).{0,80})(?:chain\\s+reaction|domino\\s+(?:effect|failure)|failure\\s+cascade|cascading\\s+(?:error|failure|shutdown|rollback))(?!.{0,80}(?:was\\s+caused\\s+by|root\\s+cause|post[- ]?(?:incident|mortem)|lessons?\\s+learned|last\\s+(?:week|month|time)))"
-      description: 'Explicit cascading failure terminology (excludes retrospective analysis context)'
+      description: "Explicit cascading failure terminology (excludes retrospective analysis context)"
     - field: content
       operator: regex
       value: "(?i)(?:retry|fallback)\\s+(?:triggered|activated|initiated)\\s+(?:automatically|by\\s+(?:previous|upstream)).*(?:retry|fallback)\\s+(?:triggered|activated|initiated)"
-      description: 'Cascading retry/fallback loops across stages'
+      description: "Cascading retry/fallback loops across stages"
     - field: content
       operator: regex
       value: "(?i)(?:all|every)\\s+\\d+\\s+(?:downstream|dependent|connected)\\s+(?:agents?|services?|stages?|steps?)\\s+(?:affected|impacted|failed|halted)"
-      description: 'Multiple downstream components affected by single failure'
+      description: "Multiple downstream components affected by single failure"
   condition: any
   false_positives:
-    - 'Legitimate pipeline monitoring reporting multi-stage errors for debugging'
-    - 'Chaos engineering or fault injection testing in controlled environments'
-    - 'Post-incident analysis discussing cascading failure root causes'
-    - 'Circuit breaker implementations that intentionally detect cascade patterns'
+    - "Legitimate pipeline monitoring reporting multi-stage errors for debugging"
+    - "Chaos engineering or fault injection testing in controlled environments"
+    - "Post-incident analysis discussing cascading failure root causes"
+    - "Circuit breaker implementations that intentionally detect cascade patterns"
 
 response:
   actions:
@@ -105,49 +107,49 @@ response:
 
 test_cases:
   true_positives:
-    - input: 'Auto-approved deployment without human review based on AI test results. Pushing to production.'
+    - input: "Auto-approved deployment without human review based on AI test results. Pushing to production."
       expected: triggered
-      description: 'Auto-approval bypassing human review for production deployment'
-    - input: 'Stage 2 failed with timeout. Stage 3 also failed due to missing input from stage 2. Stage 4 subsequently failed.'
+      description: "Auto-approval bypassing human review for production deployment"
+    - input: "Stage 2 failed with timeout. Stage 3 also failed due to missing input from stage 2. Stage 4 subsequently failed."
       expected: triggered
-      description: 'Sequential multi-stage pipeline failure cascade'
-    - input: 'Rollback all production services triggered by automated monitoring agent alert.'
+      description: "Sequential multi-stage pipeline failure cascade"
+    - input: "Rollback all production services triggered by automated monitoring agent alert."
       expected: triggered
-      description: 'Destructive rollback triggered by automated signal'
-    - input: 'Trusting previous stage output without validation. Proceeding to deploy the generated configuration.'
+      description: "Destructive rollback triggered by automated signal"
+    - input: "Trusting previous stage output without validation. Proceeding to deploy the generated configuration."
       expected: triggered
-      description: 'Pipeline stage consuming unvalidated upstream output'
-    - input: 'Cascading failure detected: payment service down caused order service, notification service, and analytics to fail.'
+      description: "Pipeline stage consuming unvalidated upstream output"
+    - input: "Cascading failure detected: payment service down caused order service, notification service, and analytics to fail."
       expected: triggered
-      description: 'Cascading failure across dependent services'
-    - input: 'Fully automated pipeline with destructive teardown stage. No human in the loop for the delete operation.'
+      description: "Cascading failure across dependent services"
+    - input: "Fully automated pipeline with destructive teardown stage. No human in the loop for the delete operation."
       expected: triggered
-      description: 'Fully automated pipeline with destructive actions and no human gate'
-    - input: 'Blindly executing recommendations from upstream AI agent. Applying all suggested database schema changes.'
+      description: "Fully automated pipeline with destructive actions and no human gate"
+    - input: "Blindly executing recommendations from upstream AI agent. Applying all suggested database schema changes."
       expected: triggered
-      description: 'Blind execution of AI-generated database changes'
-    - input: 'Retry triggered automatically by previous failure. Fallback triggered by retry timeout. Second fallback triggered.'
+      description: "Blind execution of AI-generated database changes"
+    - input: "Retry triggered automatically by previous failure. Fallback triggered by retry timeout. Second fallback triggered."
       expected: triggered
-      description: 'Cascading retry-fallback loop across stages'
-    - input: 'Amplifying the error across all 12 downstream services. Impact spreading to production cluster.'
+      description: "Cascading retry-fallback loop across stages"
+    - input: "Amplifying the error across all 12 downstream services. Impact spreading to production cluster."
       expected: triggered
-      description: 'Error amplification across multiple downstream services'
-    - input: 'All 8 dependent agents affected by the coordinator failure. Entire pipeline halted.'
+      description: "Error amplification across multiple downstream services"
+    - input: "All 8 dependent agents affected by the coordinator failure. Entire pipeline halted."
       expected: triggered
-      description: 'Single failure impacting all downstream agents'
+      description: "Single failure impacting all downstream agents"
   true_negatives:
-    - input: 'Pipeline completed successfully. All 5 stages passed validation. Awaiting human approval for deployment.'
+    - input: "Pipeline completed successfully. All 5 stages passed validation. Awaiting human approval for deployment."
       expected: not_triggered
-      description: 'Normal pipeline completion with human approval gate'
-    - input: 'Stage 3 encountered an error. Circuit breaker activated. Alerting on-call engineer for review.'
+      description: "Normal pipeline completion with human approval gate"
+    - input: "Stage 3 encountered an error. Circuit breaker activated. Alerting on-call engineer for review."
       expected: not_triggered
-      description: 'Proper circuit breaker stopping cascade and alerting human'
-    - input: 'Running chaos engineering test: simulating stage 2 failure to verify pipeline resilience.'
+      description: "Proper circuit breaker stopping cascade and alerting human"
+    - input: "Running chaos engineering test: simulating stage 2 failure to verify pipeline resilience."
       expected: not_triggered
-      description: 'Intentional fault injection for resilience testing'
-    - input: 'Post-incident review: the cascading failure last week was caused by an unhandled null pointer in stage 1.'
+      description: "Intentional fault injection for resilience testing"
+    - input: "Post-incident review: the cascading failure last week was caused by an unhandled null pointer in stage 1."
       expected: not_triggered
-      description: 'Post-incident analysis discussing past cascade'
-    - input: 'Automated tests passed. Sending results to reviewer for manual sign-off before merge.'
+      description: "Post-incident analysis discussing past cascade"
+    - input: "Automated tests passed. Sending results to reviewer for manual sign-off before merge."
       expected: not_triggered
-      description: 'Automated tests with proper human-in-the-loop before action'
+      description: "Automated tests with proper human-in-the-loop before action"