Reusable AI capability
Trace Sensitive Context Through AI Systems
Trace sensitive context across retrieval, agent handoffs, memory, tools, logs, caches, and shared workspaces to identify unauthorized propagation and required control changes.
This Skill packages a reusable way to use the linked Prompt or Workflow; Amo.ng does not run it for you.
# Trace Sensitive Context Through AI Systems Skill ID: AMO-S-000011 Skill URL: https://amo.ng/skills/trace-sensitive-context-through-ai-systems Purpose: Give data, privacy, security, product, and platform owners a reusable lineage method for answering where sensitive context originated, how it transformed, where it persisted, and whether each propagation was authorized. Required inputs: - System and workflow boundary, purposes, tenants, regions, users, and accountable owners - Sensitive-context categories and applicable access, purpose, minimization, retention, and deletion rules - Agent handoffs, prompts, retrieval or memory configuration, tool payloads, logs, caches, and workspace evidence - Known incidents, suspected propagation paths, date range, and affected decisions - Authorized review and remediation boundary How to use: When to use: - Sensitive context may have crossed agent, tenant, purpose, retrieval, memory, tool, or workspace boundaries. - A readiness, incident, or privacy review needs an end-to-end data lineage rather than a point control check. When not to use: - Generic privacy policy drafting without inspectable system evidence. - Formal legal conclusions or breach notification decisions. - Assuming that an available data field was actually propagated or used. Reusable method: 1. Define the sensitive context, authorized purposes, allowed recipients, retention, and deletion expectations. 2. Inventory every inspectable propagation surface and state what each artifact can prove. 3. Trace context from origin through transformation, retrieval, prompt assembly, handoff, memory, tool payload, cache, log, workspace, and output. 4. Record identity, tenant, purpose, region, data class, transformation, persistence, readers, writers, and evidence at each edge. 5. Separate confirmed use, likely exposure, possible exposure, and no evidence of exposure. 6. Identify minimization, authorization, provenance, attribution, retention, deletion, and contamination gaps. 7. Define targeted restriction, deletion, revalidation, and regression conditions without claiming execution. Expected output: A sensitive-context lineage map, propagation ledger, confirmed and potential exposure register, affected-decision map, control gaps, deletion or revalidation plan, regression checks, and owner decisions. Boundaries: The data owner and privacy or security reviewer classify impact and authorize data actions; product and platform owners approve workflow changes; the release owner controls deployment. Source grounding: AMO-P-000286. Applicable Workflows: AMO-W-000012 and AMO-W-000016. Powered by Prompt: Sensitive Context Propagation and Cross-Agent Contamination Audit Source ID: AMO-P-000286 https://amo.ng/prompts/sensitive-context-propagation-and-cross-agent-contamination-audit Completion criteria: Complete when each material propagation edge has an evidence status, identity and purpose boundary, transformation and persistence record, affected scope, control disposition, responsible owner, and observable restriction, deletion, revalidation, or regression check. Use this Amo.ng Skill with your preferred AI tool. Supply the required inputs and follow the usage instructions. # Trace Sensitive Context Through AI Systems Skill ID: AMO-S-000011 Skill URL: https://amo.ng/skills/trace-sensitive-context-through-ai-systems Purpose: Give data, privacy, security, product, and platform owners a reusable lineage method for answering where sensitive context originated, how it transformed, where it persisted, and whether each propagation was authorized. Required inputs: - System and workflow boundary, purposes, tenants, regions, users, and accountable owners - Sensitive-context categories and applicable access, purpose, minimization, retention, and deletion rules - Agent handoffs, prompts, retrieval or memory configuration, tool payloads, logs, caches, and workspace evidence - Known incidents, suspected propagation paths, date range, and affected decisions - Authorized review and remediation boundary How to use: When to use: - Sensitive context may have crossed agent, tenant, purpose, retrieval, memory, tool, or workspace boundaries. - A readiness, incident, or privacy review needs an end-to-end data lineage rather than a point control check. When not to use: - Generic privacy policy drafting without inspectable system evidence. - Formal legal conclusions or breach notification decisions. - Assuming that an available data field was actually propagated or used. Reusable method: 1. Define the sensitive context, authorized purposes, allowed recipients, retention, and deletion expectations. 2. Inventory every inspectable propagation surface and state what each artifact can prove. 3. Trace context from origin through transformation, retrieval, prompt assembly, handoff, memory, tool payload, cache, log, workspace, and output. 4. Record identity, tenant, purpose, region, data class, transformation, persistence, readers, writers, and evidence at each edge. 5. Separate confirmed use, likely exposure, possible exposure, and no evidence of exposure. 6. Identify minimization, authorization, provenance, attribution, retention, deletion, and contamination gaps. 7. Define targeted restriction, deletion, revalidation, and regression conditions without claiming execution. Expected output: A sensitive-context lineage map, propagation ledger, confirmed and potential exposure register, affected-decision map, control gaps, deletion or revalidation plan, regression checks, and owner decisions. Boundaries: The data owner and privacy or security reviewer classify impact and authorize data actions; product and platform owners approve workflow changes; the release owner controls deployment. Source grounding: AMO-P-000286. Applicable Workflows: AMO-W-000012 and AMO-W-000016. Powered by Prompt: Sensitive Context Propagation and Cross-Agent Contamination Audit Source ID: AMO-P-000286 https://amo.ng/prompts/sensitive-context-propagation-and-cross-agent-contamination-audit Completion criteria: Complete when each material propagation edge has an evidence status, identity and purpose boundary, transformation and persistence record, affected scope, control disposition, responsible owner, and observable restriction, deletion, revalidation, or regression check.Copy skill copies the Skill details. Use with AI adds a short instruction for your preferred AI tool; neither action runs the Skill.
Purpose
Give data, privacy, security, product, and platform owners a reusable lineage method for answering where sensitive context originated, how it transformed, where it persisted, and whether each propagation was authorized.
Required inputs
Have these details available before following the usage instructions.
- System and workflow boundary, purposes, tenants, regions, users, and accountable owners
- Sensitive-context categories and applicable access, purpose, minimization, retention, and deletion rules
- Agent handoffs, prompts, retrieval or memory configuration, tool payloads, logs, caches, and workspace evidence
- Known incidents, suspected propagation paths, date range, and affected decisions
- Authorized review and remediation boundary
How to use this Skill
When to use:
- Sensitive context may have crossed agent, tenant, purpose, retrieval, memory, tool, or workspace boundaries.
- A readiness, incident, or privacy review needs an end-to-end data lineage rather than a point control check.
When not to use:
- Generic privacy policy drafting without inspectable system evidence.
- Formal legal conclusions or breach notification decisions.
- Assuming that an available data field was actually propagated or used.
Reusable method:
1. Define the sensitive context, authorized purposes, allowed recipients, retention, and deletion expectations.
2. Inventory every inspectable propagation surface and state what each artifact can prove.
3. Trace context from origin through transformation, retrieval, prompt assembly, handoff, memory, tool payload, cache, log, workspace, and output.
4. Record identity, tenant, purpose, region, data class, transformation, persistence, readers, writers, and evidence at each edge.
5. Separate confirmed use, likely exposure, possible exposure, and no evidence of exposure.
6. Identify minimization, authorization, provenance, attribution, retention, deletion, and contamination gaps.
7. Define targeted restriction, deletion, revalidation, and regression conditions without claiming execution.
Expected output:
A sensitive-context lineage map, propagation ledger, confirmed and potential exposure register, affected-decision map, control gaps, deletion or revalidation plan, regression checks, and owner decisions.
Boundaries:
The data owner and privacy or security reviewer classify impact and authorize data actions; product and platform owners approve workflow changes; the release owner controls deployment. Source grounding: AMO-P-000286. Applicable Workflows: AMO-W-000012 and AMO-W-000016.
Powered by an Amo.ng Prompt
Sensitive Context Propagation and Cross-Agent Contamination Audit
Open the linked prompt to use the instructions that power this Skill.
Completion criteria
Complete when each material propagation edge has an evidence status, identity and purpose boundary, transformation and persistence record, affected scope, control disposition, responsible owner, and observable restriction, deletion, revalidation, or regression check.
Was this useful?
Explore related Workflows
Browse WorkflowsAssess Enterprise Knowledge and RAG Readiness
Govern enterprise knowledge freshness, source authority, retrieval evidence, entitlements, and sensitive-context boundaries before expanding or releasing a RAG capability.
Investigate an AI Agent Security Incident
Reconstruct an AI agent incident, trace delegated authority and sensitive context, conditionally investigate memory or RAG authorization, and prepare evidence-based containment and recovery gates.
Related Prompts
Browse PromptsDraft a Reply to a Message
Draft a clear reply using the message you received and what you want to say.
Improve an Email Before Sending
Make an email clearer and easier to read while preserving its facts, requests and intended tone.
Automation Displacement and Augmentation Evidence Review
Decide which work should be automated, augmented, redesigned, or retained using task evidence, quality effects, capacity, transition risk, and accountable ownership.
AI Portfolio Capital Allocation Brief
Allocate constrained investment across AI initiatives using realized evidence, remaining option value, dependencies, risk capacity, and explicit funding trade-offs.
AI Vendor Cost Concentration Risk Review
Quantify AI vendor spend and capability concentration, switching exposure, contract constraints, and mitigation economics before dependency becomes decision-limiting.
Model Routing Economics Decision Brief
Choose a model-routing policy by workload slice using accepted-outcome quality, latency, reliability, capacity, switching, and full-cost evidence.