arckit-cli 6.1.1__py3-none-any.whl

arckit_cli-6.1.1.data/data/share/arckit/.arckit/templates/agent-inventory-template.md

@@ -0,0 +1,108 @@
+---
+title: "AI Agent Inventory"
+docType: AAGI
+templateVersion: "1.0"
+---
+
+# AI Agent Inventory
+
+## Document Control
+
+| Field | Value |
+|-------|-------|
+| Document ID | ARC-{P}-AAGI-v{VERSION} |
+| Document Type | AAGI — AI Agent Inventory |
+| Project | {PROJECT_NAME} ({PROJECT_ID}) |
+| Classification | {CLASSIFICATION} |
+| Status | {STATUS} |
+| Version | {VERSION} |
+| Created | {DATE} |
+| Last Modified | {DATE} |
+| Review Cycle | Monthly |
+| Next Review | {NEXT_REVIEW_DATE} |
+| Owner | {OWNER_NAME_AND_ROLE} |
+| Reviewed By | — |
+| Approved By | — |
+| Distribution | Architecture Team, AI Programme Board |
+
+---
+
+## 1. Agent Register
+
+| Agent ID | Name | Purpose | Model | Deployment | Owner | Risk Level | Oversight |
+|----------|------|---------|-------|-----------|-------|------------|-----------|
+| AGT-001 | [Agent Name] | [Brief purpose description] | [GPT-4/Claude/etc] | [Prod/Staging/Dev/Planned] | [Owner Name/Role] | [Critical/High/Med/Low] | [Human-in-the-loop/on-the-loop/out-of-loop] |
+| AGT-002 | [Agent Name] | [Brief purpose description] | [Model] | [Env] | [Owner] | [Risk] | [Oversight] |
+| AGT-003 | [Agent Name] | [Brief purpose description] | [Model] | [Env] | [Owner] | [Risk] | [Oversight] |
+
+> **Minimum**: At least 3 agents must be documented. If fewer exist, ask user to confirm scope.
+
+## 2. Capability Matrix
+
+| Agent ID | Tools | Skills | Memory | Output Types |
+|----------|-------|--------|--------|-------------|
+| AGT-001 | [tool1, tool2, MCP client] | [skill1, skill2] | [Session/Durable/Vector/Episodic] | [Text/API/File/Action] |
+| AGT-002 | [tool1, tool2] | [skill1] | [Session/Vector] | [Text/API] |
+| AGT-003 | [tool1, tool2, tool3] | [skill1, skill2] | [Durable/Vector] | [Text/File] |
+
+## 3. Agent Dependencies
+
+```mermaid
+flowchart TD
+    AGT1[AGT-001: Orchestrator] --> AGT2[AGT-002: Research]
+    AGT1 --> AGT3[AGT-003: Analysis]
+    AGT2 --> AGT3
+```
+
+> **Note**: Update agent IDs and labels to match actual agents. Show all agent-to-agent communication paths, including data flows and orchestration chains.
+
+## 4. Security Classification
+
+| Agent ID | Data Sensitivity | Access Level | Isolation | Audit Required |
+|----------|----------------|-------------|-----------|----------------|
+| AGT-001 | [High/Med/Low] | [Full/Read/None] | [Sandboxed/Container/Native] | [Yes/No] |
+| AGT-002 | [High/Med/Low] | [Full/Read/None] | [Sandboxed/Container/Native] | [Yes/No] |
+| AGT-003 | [High/Med/Low] | [Full/Read/None] | [Sandboxed/Container/Native] | [Yes/No] |
+
+## 5. Agent Lifecycle
+
+| Agent ID | Status | Created | Last Updated | Version |
+|----------|--------|---------|-------------|---------|
+| AGT-001 | [Active/In Development/Deprecating/Retired] | [YYYY-MM-DD] | [YYYY-MM-DD] | [vX.Y] |
+| AGT-002 | [Active/In Development/Deprecating/Retired] | [YYYY-MM-DD] | [YYYY-MM-DD] | [vX.Y] |
+| AGT-003 | [Active/In Development/Deprecating/Retired] | [YYYY-MM-DD] | [YYYY-MM-DD] | [vX.Y] |
+
+## 6. Oversight Model
+
+| Agent ID | Oversight Level | Approval Gates | Monitoring | Escalation |
+|----------|----------------|---------------|------------|------------|
+| AGT-001 | [Human-in-the-loop/on-the-loop/out-of-loop] | [Gate descriptions] | [Dashboard/Alert/Audit] | [Criteria] |
+| AGT-002 | [Human-in-the-loop/on-the-loop/out-of-loop] | [Gate descriptions] | [Dashboard/Alert/Audit] | [Criteria] |
+| AGT-003 | [Human-in-the-loop/on-the-loop/out-of-loop] | [Gate descriptions] | [Dashboard/Alert/Audit] | [Criteria] |
+
+## 7. Traceability
+
+| Source | Reference | Link |
+|--------|-----------|------|
+| ADMP | [ADM Preliminary / Architecture Vision] | `ARC-{P}-ADMP-v{VER}` |
+| APP | [Application Inventory] | `ARC-{P}-APP-v{VER}` |
+| PRIN | [Principles] | `ARC-000-PRIN-v{VER}` |
+| AAGR | [Agent Architecture Specification] | `ARC-{P}-AAGR-v{VER}` |
+| AAOV | [Agent Governance Framework] | `ARC-{P}-AAOV-v{VER}` |
+| AASE | [Agent Security Architecture] | `ARC-{P}-AASE-v{VER}` |
+
+---
+
+## Revision History
+
+| Version | Date | Author | Changes | Approved By | Approval Date |
+|---------|------|--------|---------|-------------|--------------|
+| 1.0 | {DATE} | ArcKit AI | Initial creation from `/arckit:agent-inventory` command | — | — |
+
+---
+
+**Generated by**: ArcKit `/arckit:agent-inventory` command
+**Generated on**: {DATE}
+**ArcKit Version**: {ARCKIT_VERSION}
+**Project**: {PROJECT_NAME} (Project {PROJECT_ID})
+**AI Model**: [Model Name]

arckit_cli-6.1.1.data/data/share/arckit/.arckit/templates/agent-maturity-template.md

@@ -0,0 +1,388 @@
+---
+title: "Agent Program Maturity Model"
+docType: AAMT
+templateVersion: "1.0"
+---
+
+# Agent Program Maturity Model
+
+## Document Control
+
+<!-- DOC-CONTROL-HEADER -->
+<!-- Resolved at command-execution time to _partials/document-control-uk.md or _partials/document-control-uae.md based on plugin userConfig classification_scheme + governance_framework. See _partials/RENDERING.md (when present). -->
+
+## Revision History
+
+| Version | Date | Author | Changes | Approved By | Approval Date |
+|---------|------|--------|---------|-------------|---------------|
+| [VERSION] | [DATE] | ArcKit AI | Initial creation from `/arckit:agent-maturity` command | [PENDING] | [PENDING] |
+
+---
+
+## 1. Maturity Model Framework
+
+### 1.1 Framework Overview
+
+This maturity model assesses the AI agent program across five dimensions at five levels of maturity. Each level represents a progressive stage of capability development.
+
+| Level | Name | Description |
+|-------|------|-------------|
+| L1 | Ad-hoc | No formal processes, reactive, undocumented, person-dependent |
+| L2 | Reactive | Processes defined post-incident, some documentation, inconsistent |
+| L3 | Defined | Standard processes documented, proactive management, measured outcomes |
+| L4 | Managed | Metrics-driven, data-led decisions, continuous improvement |
+| L5 | Optimized | Continuous improvement, predictive, industry-leading, automated |
+
+### 1.2 Level Definitions by Dimension
+
+#### Design
+
+| Level | Characteristics |
+|-------|----------------|
+| L1 | Designs created ad-hoc, no reuse, decisions undocumented |
+| L2 | Basic designs documented after issues arise, some patterns identified |
+| L3 | Standard design patterns documented, reviews required, reusable templates |
+| L4 | Design quality measured, automated validation, architecture decision records |
+| L5 | AI-assisted design optimisation, predictive pattern matching, industry contribution |
+
+#### Governance
+
+| Level | Characteristics |
+|-------|----------------|
+| L1 | No formal governance, decisions made by individuals |
+| L2 | Basic oversight after incidents, ad-hoc reviews |
+| L3 | Governance framework established, regular reviews, documented authority |
+| L4 | Automated compliance monitoring, governance metrics, audit trails |
+| L5 | Self-governing framework, predictive compliance, industry standards contribution |
+
+#### Security
+
+| Level | Characteristics |
+|-------|----------------|
+| L1 | No formal security design, reactive to breaches |
+| L2 | Basic security controls added after incidents |
+| L3 | Security by design, threat modelling, regular security testing |
+| L4 | Automated security monitoring, continuous vulnerability scanning, security metrics |
+| L5 | Predictive threat detection, zero-trust architecture, security automation |
+
+#### Integration
+
+| Level | Characteristics |
+|-------|----------------|
+| L1 | Point-to-point integrations, no standards, manual connections |
+| L2 | Basic integration patterns, some API documentation |
+| L3 | Standardised integration patterns, API specifications, integration testing |
+| L4 | API management platform, automated integration testing, observability |
+| L5 | Self-healing integrations, automated discovery, integration orchestration |
+
+#### Operations
+
+| Level | Characteristics |
+|-------|----------------|
+| L1 | Manual operations, no monitoring, reactive to failures |
+| L2 | Basic monitoring, runbooks created after incidents |
+| L3 | Standard operating procedures, alerting, SLA tracking |
+| L4 | Automated operations, predictive monitoring, MTTR targets |
+| L5 | Self-healing operations, predictive scaling, chaos engineering |
+
+---
+
+## 2. Current State Assessment
+
+### 2.1 Assessment Matrix
+
+| Dimension | Level | Evidence | Gaps |
+|-----------|-------|----------|------|
+| Design | [Current level, e.g., L2] | [Concrete evidence — e.g., "Architecture decisions documented in 2 of 5 agents, no reusable patterns"] | [Gaps — e.g., "No design review process, limited pattern reuse across agents"] |
+| Governance | [Current level, e.g., L1] | [Concrete evidence — e.g., "No formal governance board for AI agents, ad-hoc oversight"] | [Gaps — e.g., "No audit trail, no compliance monitoring, no authority matrix"] |
+| Security | [Current level, e.g., L2] | [Concrete evidence — e.g., "Guardrails defined in 3 agents, basic prompt injection checks"] | [Gaps — e.g., "No threat model, no automated security testing, no security metrics"] |
+| Integration | [Current level, e.g., L1] | [Concrete evidence — e.g., "Point-to-point API connections, no integration standards"] | [Gaps — e.g., "No API management, no integration testing, no observability"] |
+| Operations | [Current level, e.g., L2] | [Concrete evidence — e.g., "Basic monitoring via cloud provider, ad-hoc runbooks"] | [Gaps — e.g., "No SLA tracking, no alerting thresholds, no incident response plan"] |
+
+### 2.2 Maturity Heatmap
+
+```mermaid
+quadrantChart
+    title Agent Program Maturity — Current State
+    x-axis__Low --> High
+    y-axis__High Priority --> Low Priority
+    quadrant-1 Quick Wins
+    quadrant-2 Monitor
+    quadrant-3 Strategic Investments
+    quadrant-4 Develop
+    "Design", [0.3, 0.4]
+    "Governance", [0.1, 0.3]
+    "Security", [0.2, 0.5]
+    "Integration", [0.2, 0.2]
+    "Operations", [0.3, 0.6]
+```
+
+### 2.3 Dimension Analysis
+
+#### Design — [Current Level]
+
+**Strengths**:
+
+- [Strength 1]
+- [Strength 2]
+
+**Areas for improvement**:
+
+- [Gap 1 — what needs to change]
+- [Gap 2 — specific artefact or process missing]
+
+**Evidence summary**:
+
+- [Artefact reference 1]
+- [Artefact reference 2]
+
+#### Governance — [Current Level]
+
+**Strengths**:
+
+- [Strength 1]
+- [Strength 2]
+
+**Areas for improvement**:
+
+- [Gap 1]
+- [Gap 2]
+
+**Evidence summary**:
+
+- [Artefact reference 1]
+- [Artefact reference 2]
+
+#### Security — [Current Level]
+
+**Strengths**:
+
+- [Strength 1]
+- [Strength 2]
+
+**Areas for improvement**:
+
+- [Gap 1]
+- [Gap 2]
+
+**Evidence summary**:
+
+- [Artefact reference 1]
+- [Artefact reference 2]
+
+#### Integration — [Current Level]
+
+**Strengths**:
+
+- [Strength 1]
+- [Strength 2]
+
+**Areas for improvement**:
+
+- [Gap 1]
+- [Gap 2]
+
+**Evidence summary**:
+
+- [Artefact reference 1]
+- [Artefact reference 2]
+
+#### Operations — [Current Level]
+
+**Strengths**:
+
+- [Strength 1]
+- [Strength 2]
+
+**Areas for improvement**:
+
+- [Gap 1]
+- [Gap 2]
+
+**Evidence summary**:
+
+- [Artefact reference 1]
+- [Artefact reference 2]
+
+---
+
+## 3. Target State
+
+### 3.1 Target Levels
+
+| Dimension | Current | Target | Gap | Rationale |
+|-----------|---------|--------|-----|-----------|
+| Design | [L2] | [L4] | [+2] | [e.g., "Support multi-agent programmes with standardised patterns and automated design validation"] |
+| Governance | [L1] | [L3] | [+2] | [e.g., "Regulatory compliance requires documented governance framework within 6 months"] |
+| Security | [L2] | [L4] | [+2] | [e.g., "Security incidents require proactive threat modelling and automated security testing"] |
+| Integration | [L1] | [L3] | [+2] | [e.g., "Enterprise integration requires standardised APIs and integration testing"] |
+| Operations | [L2] | [L4] | [+2] | [e.g., "Production operations require SLA tracking, alerting, and automated response"] |
+
+### 3.2 Priority Ranking
+
+| Rank | Dimension | Priority | Justification |
+|------|-----------|----------|---------------|
+| 1 | [Dimension] | Critical | [Why this is the highest priority] |
+| 2 | [Dimension] | High | [Why this ranks second] |
+| 3 | [Dimension] | Medium | [Why this ranks third] |
+| 4 | [Dimension] | Medium | [Why this ranks fourth] |
+| 5 | [Dimension] | Low | [Why this ranks last] |
+
+---
+
+## 4. Improvement Roadmap
+
+### 4.1 Improvement Initiatives
+
+| Initiative | Dimension | From | To | Timeline | Investment | Owner |
+|------------|-----------|------|----|----------|------------|-------|
+| [Initiative 1 — e.g., "Design Pattern Library"] | [Design] | [L2] | [L3] | [Q1–Q2] | [£X / FTEs] | [Team] |
+| [Initiative 2 — e.g., "Governance Board Setup"] | [Governance] | [L1] | [L2] | [Q1] | [£X / FTEs] | [Team] |
+| [Initiative 3 — e.g., "Security Testing Automation"] | [Security] | [L2] | [L3] | [Q2–Q3] | [£X / FTEs] | [Team] |
+| [Initiative 4 — e.g., "API Gateway Implementation"] | [Integration] | [L1] | [L2] | [Q2] | [£X / FTEs] | [Team] |
+| [Initiative 5 — e.g., "Operations Dashboard"] | [Operations] | [L2] | [L3] | [Q3–Q4] | [£X / FTEs] | [Team] |
+
+### 4.2 Initiative Details
+
+#### [Initiative 1]: [Name]
+
+**Dimension**: [Design]
+**Scope**: [What this initiative will deliver]
+**Activities**:
+
+- [Activity 1]
+- [Activity 2]
+- [Activity 3]
+
+**Success Criteria**:
+
+- [Criterion 1 — measurable outcome]
+- [Criterion 2 — measurable outcome]
+
+**Dependencies**: [Depends on / Enables]
+**Risks**: [Key risks and mitigations]
+
+#### [Initiative 2]: [Name]
+
+**Dimension**: [Governance]
+**Scope**: [What this initiative will deliver]
+**Activities**:
+
+- [Activity 1]
+- [Activity 2]
+- [Activity 3]
+
+**Success Criteria**:
+
+- [Criterion 1 — measurable outcome]
+- [Criterion 2 — measurable outcome]
+
+**Dependencies**: [Depends on / Enables]
+**Risks**: [Key risks and mitigations]
+
+#### [Initiative 3]: [Name]
+
+**Dimension**: [Security]
+**Scope**: [What this initiative will deliver]
+**Activities**:
+
+- [Activity 1]
+- [Activity 2]
+- [Activity 3]
+
+**Success Criteria**:
+
+- [Criterion 1 — measurable outcome]
+- [Criterion 2 — measurable outcome]
+
+**Dependencies**: [Depends on / Enables]
+**Risks**: [Key risks and mitigations]
+
+### 4.3 Timeline
+
+```mermaid
+gantt
+    title Agent Maturity Improvement Roadmap
+    dateFormat  YYYY-MM
+    section Phase 1 — Foundation
+    Governance Board Setup       :active, gov1, 2026-Q1, 3M
+    Design Pattern Library        :active, des1, 2026-Q1, 6M
+    section Phase 2 — Capability
+    Security Testing Automation   :active, sec1, 2026-Q2, 6M
+    API Gateway Implementation    :active, int1, 2026-Q2, 4M
+    section Phase 3 — Optimisation
+    Operations Dashboard          :active, ops1, 2026-Q3, 6M
+```
+
+### 4.4 Dependencies and Sequencing
+
+| Initiative | Depends On | Blocks |
+|------------|-----------|--------|
+| [Initiative 1] | [None / Initiative X] | [Initiative Y] |
+| [Initiative 2] | [None] | [Initiative 3] |
+| [Initiative 3] | [Initiative 2] | [None] |
+
+---
+
+## 5. Benchmarks
+
+### 5.1 Industry Benchmarking
+
+| Benchmark | Standard | Industry Average | Our Position | Gap |
+|-----------|----------|----------------|--------------|-----|
+| Design maturity | CMMI Software | L3 | [L2] | [-1] |
+| Governance maturity | NIST AI RMF | L3 | [L1] | [-2] |
+| Security maturity | OWASP LLM Top 10 | L3 | [L2] | [-1] |
+| Integration maturity | TOGAF | L3 | [L1] | [-2] |
+| Operations maturity | ITIL 4 | L4 | [L2] | [-2] |
+
+### 5.2 Peer Comparison
+
+| Peer Organisation | Overall Maturity | Strongest Dimension | Weakest Dimension |
+|-------------------|------------------|---------------------|--------------------|
+| [Organisation A] | [Level / Score] | [Dimension] | [Dimension] |
+| [Organisation B] | [Level / Score] | [Dimension] | [Dimension] |
+| **Our Programme** | [Level / Score] | [Dimension] | [Dimension] |
+
+### 5.3 Regulatory Alignment
+
+| Regulation | Requirement | Our Status | Evidence |
+|------------|-------------|------------|----------|
+| [Regulation 1] | [Requirement] | [Compliant/Partial/Non-compliant] | [Evidence] |
+| [Regulation 2] | [Requirement] | [Compliant/Partial/Non-compliant] | [Evidence] |
+
+---
+
+## 6. Traceability
+
+### 6.1 Source Artefacts
+
+| Source | Document ID | Contribution |
+|--------|-------------|--------------|
+| Agent Inventory | ARC-{P}-AAGI-v{VERSION} | [Agent inventory and capability baseline] |
+| Agent Architecture Specification | ARC-{P}-AAGR-v{VERSION} | [Design patterns, tool contracts, guardrails] |
+| Agent Governance Framework | ARC-{P}-AAOV-v{VERSION} | [Governance maturity evidence] |
+| Architecture Principles | ARC-000-PRIN-v{VERSION} | [Principles alignment for maturity targets] |
+
+### 6.2 Cross-References
+
+| Target | Document ID | Usage |
+|--------|-------------|-------|
+| Agent Design improvements | ARC-{P}-AAGR-v{NEXT} | [Design initiative outputs] |
+| Governance updates | ARC-{P}-AAOV-v{NEXT} | [Governance initiative outputs] |
+| Roadmap | ARC-{P}-ROAD-v{VERSION} | [Improvement roadmap integration] |
+
+### 6.3 External References
+
+| Reference | Source | Citation |
+|-----------|--------|----------|
+| [External benchmark / framework] | [Source name] | [Citation ID] |
+
+---
+
+> **Generated by**: ArcKit `/arckit:agent-maturity` command
+> **Generated on**: [DATE] [TIME] GMT
+> **ArcKit Version**: [VERSION]
+> **Project**: [PROJECT_NAME] (Project [PROJECT_ID])
+> **AI Model**: [MODEL_NAME]
+> **Generation Context**: [SOURCE_DOCUMENTS]

arckit_cli-6.1.1.data/data/share/arckit/.arckit/templates/agent-security-template.md

@@ -0,0 +1,213 @@
+---
+title: "Agent Security Architecture"
+docType: AASE
+templateVersion: "1.0"
+---
+
+# Agent Security Architecture
+
+## Document Control
+
+<!-- DOC-CONTROL-HEADER -->
+<!-- Resolved at command-execution time to _partials/document-control-uk.md or _partials/document-control-uae.md based on plugin userConfig classification_scheme + governance_framework. See _partials/RENDERING.md (when present). -->
+
+## Revision History
+
+| Version | Date | Author | Changes | Approved By | Approval Date |
+|---------|------|--------|---------|-------------|---------------|
+| [VERSION] | [DATE] | ArcKit AI | Initial creation from `/arckit:agent-security` command | [PENDING] | [PENDING] |
+
+---
+
+## 1. Threat Model
+
+### 1.1 Attack Surface
+
+```mermaid
+mindmap
+  root((Threat Model))
+    Input
+      Prompt injection
+      Data poisoning
+      Context overflow
+      Adversarial inputs
+      Model probing
+      Supply chain attacks
+    Execution
+      Tool abuse
+      Privilege escalation
+      Sandboxing escape
+      Resource exhaustion
+      Tool chain manipulation
+      Cross-agent injection
+    Output
+      Data exfiltration
+      PII leakage
+      Hallucination
+      Prompt template extraction
+      Instruction leakage
+      Model weights extraction
+    Infrastructure
+      Model theft
+      API abuse
+      Side-channel attacks
+      Denial of service
+      Credential theft
+      Infrastructure manipulation
+```
+
+### 1.2 Threat Summary
+
+| Threat Category | Dimension | Severity | Frequency | Mitigation |
+|----------------|-----------|----------|-----------|-----------|
+| [Category 1] | [Dimension] | [Critical/High/Med/Low] | [Likelihood] | [Control] |
+| [Category 2] | [Dimension] | [Critical/High/Med/Low] | [Likelihood] | [Control] |
+
+---
+
+## 2. Sandboxing Architecture
+
+| Component | Isolation | Enforcement |
+|-----------|-----------|-------------|
+| Agent process | [Container/Sandbox] | [OS-level] |
+| Tool execution | [Separate process] | [Permission boundary] |
+| Memory access | [Scoped] | [RBAC] |
+| Network access | [API boundary] | [Egress control] |
+| Data I/O | [Context sandbox] | [Input/output separation] |
+
+### 2.1 Isolation Layers
+
+| Layer | Technology | Scope | Escape Risk |
+|-------|-----------|-------|------------|
+| [OS Level] | [Container/Sandbox] | [Agent process] | [Low/Med/High] |
+| [App Level] | [Wasm/Virtualisation] | [Tool execution] | [Low/Med/High] |
+| [Data Level] | [Encryption/Tokenisation] | [Memory/Data] | [Low/Med/High] |
+
+---
+
+## 3. Tool Permission Matrix
+
+| Tool | Permission | Agent Access | Risk | Mitigation |
+|------|-----------|-------------|------|-----------|
+| [Tool 1] | [Read/Write/Execute] | [Full/Limited/None] | [Low/Med/High] | [Control] |
+| [Tool 2] | [Read/Write/Execute] | [Full/Limited/None] | [Low/Med/High] | [Control] |
+| [Tool 3] | [Read/Write/Execute] | [Full/Limited/None] | [Low/Med/High] | [Control] |
+| [Tool 4] | [Read/Write/Execute] | [Full/Limited/None] | [Low/Med/High] | [Control] |
+| [Tool 5] | [Read/Write/Execute] | [Full/Limited/None] | [Low/Med/High] | [Control] |
+
+---
+
+## 4. Data Handling Policy
+
+| Data Type | Handling | Agent Access | Retention |
+|-----------|----------|-------------|-----------|
+| PII | [Encrypt at rest, mask in prompts] | [Allowed/Blocked] | [X days] |
+| Sensitive | [Encrypted, scoped access] | [Allowed/Blocked] | [X days] |
+| Public | [Standard handling] | [Unrestricted] | [X days] |
+| Model weights | [Integrity verified, restricted] | [Read-only] | [Indefinite] |
+
+---
+
+## 5. Prompt Injection Defences
+
+| Defence | Method | Coverage |
+|---------|--------|----------|
+| Input sanitization | [Regex, LLM guard] | [Direct prompts] |
+| Context boundary | [System prompt separation] | [Indirect injection] |
+| Output validation | [Schema check, content filter] | [Response] |
+| Tool call validation | [Permission check, arg validation] | [Tool abuse] |
+| [Defence 5] | [Method] | [Coverage] |
+| [Defence 6] | [Method] | [Coverage] |
+
+---
+
+## 6. Output Validation Pipeline
+
+```mermaid
+flowchart LR
+    A[Agent Output] --> B[Schema Validation]
+    B --> C[Content Filter]
+    C --> D[PII Scanner]
+    D --> E[Human Review (if flagged)]
+    E --> F[Release]
+```
+
+### 6.1 Pipeline Details
+
+| Stage | Check | Action on Failure | Owner |
+|-------|-------|-------------------|-------|
+| Schema Validation | [Format, structure, type] | [Reject / Flag] | [System] |
+| Content Filter | [Policy, safety, compliance] | [Block / Review] | [System] |
+| PII Scanner | [Personal data detection] | [Redact / Review] | [System] |
+| Human Review | [Manual assessment] | [Approve / Reject] | [Reviewer] |
+| Release | [Final gate] | [Deliver / Escalate] | [System] |
+
+---
+
+## 7. Secret Management
+
+| Secret Type | Storage | Access | Rotation |
+|-------------|---------|--------|----------|
+| API keys | [Hashicorp Vault] | [Scoped] | [90 days] |
+| Model credentials | [Secrets manager] | [Scoped] | [90 days] |
+| Database credentials | [Vault] | [Scoped] | [30 days] |
+| Encryption keys | [HSM / KMS] | [Scoped] | [30 days] |
+
+### 7.1 Secret Lifecycle
+
+| Phase | Action | Automation | Audit |
+|-------|--------|-----------|-------|
+| Creation | [Generated / Provided] | [Automated] | [Logged] |
+| Storage | [Encrypted at rest] | [Vault-managed] | [Access log] |
+| Usage | [Scoped, time-limited] | [Injected at runtime] | [Usage log] |
+| Rotation | [Automated cycle] | [Scheduled] | [Audit trail] |
+| Revocation | [Immediate disable] | [On-demand] | [Event log] |
+
+---
+
+## 8. Incident Response
+
+| Severity | Response Time | Owner | Actions |
+|----------|---------------|-------|---------|
+| Critical | [<15 min] | [SRE] | [Isolate, investigate, patch] |
+| High | [<1h] | [Security] | [Contain, assess, remediate] |
+| Medium | [<24h] | [Team] | [Patch, monitor] |
+| Low | [<72h] | [Team] | [Log, review, improve] |
+
+### 8.1 Incident Lifecycle
+
+```mermaid
+flowchart LR
+    A[Detection] --> B[Triage]
+    B --> C{Severity}
+    C -->|Critical| D[Immediate isolation]
+    C -->|High| E[Containment]
+    C -->|Medium| F[Investigation]
+    D --> G[Remediation]
+    E --> G
+    F --> G
+    G --> H[Post-mortem]
+    H --> I[Prevention update]
+```
+
+---
+
+## 9. Traceability
+
+| Source | Reference | Link |
+|--------|-----------|------|
+| AAGI | [Agent Inventory ID] | [Agent list, risk classifications] |
+| AAGR | [Agent Design ID] | [Tool contracts, architecture] |
+| SEC / SBD | [Secure by Design ID] | [Security controls, threat models] |
+| AAOV | [Governance ID] | [Oversight tiers, compliance] |
+
+---
+
+## Generation Metadata
+
+**Generated by**: ArcKit `/arckit:agent-security` command
+**Generated on**: [DATE] [TIME] GMT
+**ArcKit Version**: [ARCKIT_VERSION]
+**Project**: [PROJECT_NAME] (Project [PROJECT_ID])
+**AI Model**: [Model name]
+**Generation Context**: [Brief note about source documents used]