@bluefly/openstandardagents 0.5.0 → 0.5.3

package/examples/showcase/PRODUCTION_EXAMPLES.md

@@ -1,499 +0,0 @@
-# Production-Ready OSSA Agent Examples
-
-This directory contains 5 comprehensive production-ready agent examples showcasing different capabilities, use cases, and patterns in OSSA v0.4.1.
-
-## Overview
-
-These examples demonstrate:
-- Different domains and use cases
-- A2A (Agent-to-Agent) communication patterns
-- Various autonomy modes and human-in-the-loop configurations
-- Token efficiency optimization strategies
-- Event-driven and scheduled execution
-- Multi-agent coordination and team leadership
-- Comprehensive observability and monitoring
-- Security and compliance patterns
-
-## Examples
-
-### 1. Compliance Checker Agent
-
-**File:** `compliance-checker.ossa.yaml`
-
-**Domain:** Governance / Compliance
-
-**Description:** Enterprise compliance checker for SOC2, HIPAA, and GDPR frameworks with policy validation, audit reporting, and regulatory compliance.
-
-**Key Features:**
-- **Frameworks:** SOC2, HIPAA, GDPR
-- **Autonomy:** High - Autonomous operation with escalation
-- **A2A Enabled:** Yes - Queries policy database, coordinates with other compliance systems
-- **Capabilities:**
-  - Policy validation against regulatory frameworks
-  - Comprehensive compliance audits
-  - Regulatory report generation
-  - Policy database queries
-- **Use Cases:**
-  - Continuous compliance monitoring
-  - Regulatory audit preparation
-  - Policy enforcement
-  - Compliance risk assessment
-
-**Notable Configuration:**
-```yaml
-autonomy:
-  mode: autonomous
-  humanApproval: false
-  escalation:
-    enabled: true
-    triggers:
-      - critical_violation
-      - framework_change
-      - audit_failure
-
-a2a:
-  enabled: true
-  capabilities:
-    - policy-validation
-    - compliance-audit
-    - regulatory-reporting
-  protocols:
-    - rest
-    - grpc
-    - kafka
-```
-
-**Resources:** 2Gi RAM, 1000m CPU, 10Gi Storage
-
----
-
-### 2. Data Processing Pipeline Agent
-
-**File:** `data-processing-pipeline.ossa.yaml`
-
-**Domain:** Data Engineering / ETL
-
-**Description:** Data processing pipeline agent with ETL, transformation, and validation capabilities using pandas, dbt, and great_expectations.
-
-**Key Features:**
-- **Tools:** pandas, dbt, great_expectations
-- **Autonomy:** Supervised - Human approval required for critical operations
-- **A2A Enabled:** No - Standalone data processing
-- **High Resource Requirements:** 8Gi RAM, 4000m CPU, 100Gi Storage
-- **Capabilities:**
-  - Data extraction from multiple sources (database, API, file, stream)
-  - Data transformation using pandas and dbt
-  - Data loading to various destinations
-  - Data quality validation with great_expectations
-- **Use Cases:**
-  - Daily batch ETL jobs
-  - Real-time data streaming
-  - Data warehouse loading
-  - Data quality monitoring
-
-**Notable Configuration:**
-```yaml
-autonomy:
-  mode: supervised
-  humanApproval: true
-  approvalRequired:
-    - before_load
-    - on_validation_failure
-    - on_schema_change
-
-scheduling:
-  enabled: true
-  cron: "0 2 * * *"  # Daily at 2 AM
-  timezone: UTC
-  maxConcurrent: 1
-
-retry:
-  enabled: true
-  maxAttempts: 3
-  backoff:
-    type: exponential
-    initialDelay: 5s
-    maxDelay: 300s
-```
-
-**Resources:** 8Gi RAM (16Gi limit), 4000m CPU (8000m limit), 100Gi Storage
-
----
-
-### 3. Team Leader Agent
-
-**File:** `team-leader.ossa.yaml`
-
-**Domain:** Orchestration / Multi-Agent Coordination
-
-**Description:** Multi-agent team leader with task distribution, coordination, and progress tracking capabilities.
-
-**Key Features:**
-- **Autonomy:** Autonomous - Full team leadership authority
-- **A2A Enabled:** Yes - Leader role with team discovery
-- **Team Discovery:** Enabled with auto-registration
-- **Capabilities:**
-  - Task distribution based on capabilities
-  - Team coordination (sequential, parallel, conditional, fan-out/fan-in)
-  - Progress tracking and reporting
-  - Team member discovery
-- **Use Cases:**
-  - Multi-agent workflow orchestration
-  - Dynamic task assignment
-  - Team workload balancing
-  - Dependency management
-
-**Notable Configuration:**
-```yaml
-autonomy:
-  mode: autonomous
-  humanApproval: false
-
-a2a:
-  enabled: true
-  role: leader
-  capabilities:
-    - task-distribution
-    - team-coordination
-    - progress-tracking
-    - agent-discovery
-  discovery:
-    enabled: true
-    broadcast: true
-    interval: 60s
-
-teamDiscovery:
-  enabled: true
-  autoRegister: true
-  heartbeat:
-    interval: 30s
-    timeout: 90s
-
-messaging:
-  transport: redis
-  events:
-    publish:
-      - task.assigned
-      - task.completed
-      - coordination.started
-      - team.member.discovered
-```
-
-**Resources:** 2Gi RAM, 2000m CPU, 5Gi Storage
-
----
-
-### 4. Security Scanner Enhanced Agent
-
-**File:** `security-scanner-enhanced.ossa.yaml`
-
-**Domain:** Security Operations
-
-**Description:** Advanced security scanner with vulnerability scanning, penetration testing, and compliance checking using nmap, OWASP ZAP, and custom scanners.
-
-**Key Features:**
-- **Tools:** nmap, OWASP ZAP, custom scanners
-- **Autonomy:** Medium - Semi-autonomous with approval gates
-- **Execution:** Event-driven (kHook triggers)
-- **Triggers:** Deployment events, code commits, configuration changes, webhooks, scheduled
-- **Capabilities:**
-  - Vulnerability scanning (network, web app, container, cloud, API)
-  - Penetration testing (authorized scope)
-  - Compliance checking (PCI-DSS, ISO27001, NIST, CIS, OWASP Top 10)
-- **Use Cases:**
-  - Continuous security scanning
-  - Pre-deployment security validation
-  - Compliance verification
-  - Threat detection
-
-**Notable Configuration:**
-```yaml
-autonomy:
-  mode: semi-autonomous
-  humanApproval: true
-  approvalRequired:
-    - before_penetration_test
-    - before_production_scan
-    - on_critical_finding
-
-triggers:
-  khook:
-    enabled: true
-    events:
-      - type: deployment
-        action: trigger_scan
-        filter:
-          environment: [production, staging]
-      - type: code_commit
-        action: trigger_scan
-  webhook:
-    enabled: true
-  schedule:
-    enabled: true
-    cron: "0 0 * * 0"  # Weekly
-
-observability:
-  alerts:
-    - name: critical_vulnerability_found
-      condition: vulnerabilities_found{severity="critical"} > 0
-      notification: pagerduty
-```
-
-**Resources:** 4Gi RAM, 2000m CPU, 20Gi Storage
-
----
-
-### 5. Customer Support Agent
-
-**File:** `customer-support.ossa.yaml`
-
-**Domain:** Customer Service
-
-**Description:** Intelligent customer support agent with ticket triage, auto-response, escalation, and 70-95% token efficiency optimization.
-
-**Key Features:**
-- **LLM:** Claude 3.5 Sonnet with prompt caching
-- **Token Efficiency:** 70-95% savings through aggressive caching
-- **Autonomy:** Collaborative - Human-in-the-loop for complex cases
-- **Capabilities:**
-  - Ticket triage with sentiment analysis
-  - Automated response generation
-  - Smart escalation management
-  - Knowledge base integration
-- **Use Cases:**
-  - First-line customer support
-  - Ticket categorization and routing
-  - FAQ responses
-  - Escalation to human agents
-
-**Notable Configuration:**
-```yaml
-llm:
-  provider: anthropic
-  model: claude-3-5-sonnet-20241022
-  temperature: 0.7
-  maxTokens: 8192
-  features:
-    promptCaching: true
-    tokenEfficiency: true
-    streamingResponse: true
-  caching:
-    enabled: true
-    strategy: aggressive
-    cacheableElements:
-      - system_prompt
-      - knowledge_base_context
-      - company_policies
-      - common_faqs
-    ttl: 300  # 5 minutes
-    targetSavings: 0.85  # Target 85% token savings
-
-autonomy:
-  mode: collaborative
-  humanInTheLoop:
-    enabled: true
-    triggers:
-      - escalation_requested
-      - low_confidence
-      - negative_sentiment
-
-sla:
-  responseTime:
-    urgent: 15m
-    high: 1h
-    medium: 4h
-    low: 24h
-```
-
-**Resources:** 2Gi RAM, 1000m CPU, 5Gi Storage
-
----
-
-## Comparison Matrix
-
-| Agent | Domain | Autonomy | A2A | Token Efficiency | Event-Driven | Resources |
-|-------|--------|----------|-----|------------------|--------------|-----------|
-| Compliance Checker | Governance | High | Yes | Yes (caching) | No | Medium |
-| Data Processing Pipeline | Data Engineering | Supervised | No | No | No | High |
-| Team Leader | Orchestration | Autonomous | Yes (Leader) | Yes (caching) | No | Medium |
-| Security Scanner | Security | Medium | No | Yes (caching) | Yes (kHook) | Medium-High |
-| Customer Support | Support | Collaborative | No | Yes (85% savings) | No | Medium |
-
-## Pattern Demonstrations
-
-### 1. A2A Communication Patterns
-
-**Compliance Checker** and **Team Leader** demonstrate different A2A patterns:
-- Compliance Checker: Query-based A2A (policy database)
-- Team Leader: Coordination-based A2A (task distribution, team discovery)
-
-### 2. Autonomy Modes
-
-- **Autonomous:** Team Leader, Compliance Checker
-- **Semi-Autonomous:** Security Scanner
-- **Supervised:** Data Processing Pipeline
-- **Collaborative:** Customer Support
-
-### 3. Human-in-the-Loop
-
-Different approval strategies:
-- **Pre-operation approval:** Security Scanner (before pen tests)
-- **Conditional approval:** Data Processing Pipeline (on validation failure)
-- **Escalation-based:** Customer Support (low confidence triggers)
-
-### 4. Token Efficiency
-
-Three approaches to token optimization:
-- **Aggressive caching:** Customer Support (85% target savings)
-- **Standard caching:** Compliance Checker, Team Leader, Security Scanner
-- **No optimization:** Data Processing Pipeline (non-LLM intensive)
-
-### 5. Event-Driven Execution
-
-**Security Scanner** shows comprehensive trigger configuration:
-- kHook events (deployment, code commit, configuration change)
-- Webhook endpoints
-- Scheduled execution (cron)
-
-### 6. Resource Allocation
-
-Different resource profiles for different workloads:
-- **High:** Data Processing Pipeline (8-16Gi RAM, 4-8 CPU cores)
-- **Medium:** Security Scanner (4Gi RAM, 2 CPU cores)
-- **Light:** Compliance Checker, Team Leader, Customer Support (2Gi RAM, 1-2 CPU cores)
-
-## Usage
-
-### Validation
-
-Validate all manifests:
-
-```bash
-# Individual validation
-npm run validate:manifest -- examples/showcase/compliance-checker.ossa.yaml
-npm run validate:manifest -- examples/showcase/data-processing-pipeline.ossa.yaml
-npm run validate:manifest -- examples/showcase/team-leader.ossa.yaml
-npm run validate:manifest -- examples/showcase/security-scanner-enhanced.ossa.yaml
-npm run validate:manifest -- examples/showcase/customer-support.ossa.yaml
-
-# Validate all examples
-npm run validate:examples
-```
-
-### Export to Platforms
-
-Export to different platforms (LangChain, CrewAI, AutoGen, etc.):
-
-```bash
-# Export Compliance Checker to LangChain
-npx ossa export examples/showcase/compliance-checker.ossa.yaml --output langchain --output-dir ./exported/compliance-checker
-
-# Export Team Leader to CrewAI
-npx ossa export examples/showcase/team-leader.ossa.yaml --output crewai --output-dir ./exported/team-leader
-
-# Export Security Scanner to Kubernetes
-npx ossa export examples/showcase/security-scanner-enhanced.ossa.yaml --output kubernetes --output-dir ./exported/security-scanner
-
-# Export all to multiple platforms
-for agent in compliance-checker data-processing-pipeline team-leader security-scanner-enhanced customer-support; do
-  npx ossa export examples/showcase/$agent.ossa.yaml --output langchain --output-dir ./exported/$agent/langchain
-  npx ossa export examples/showcase/$agent.ossa.yaml --output crewai --output-dir ./exported/$agent/crewai
-  npx ossa export examples/showcase/$agent.ossa.yaml --output kubernetes --output-dir ./exported/$agent/k8s
-done
-```
-
-### Deployment
-
-Deploy to Kubernetes:
-
-```bash
-# Generate Kubernetes manifests
-npx ossa export examples/showcase/compliance-checker.ossa.yaml --output kubernetes --output-dir ./k8s
-
-# Apply to cluster
-kubectl apply -f ./k8s/
-```
-
-### View Agent Information
-
-```bash
-# View detailed agent information
-npx ossa info examples/showcase/compliance-checker.ossa.yaml
-npx ossa info examples/showcase/team-leader.ossa.yaml
-```
-
-## Best Practices Demonstrated
-
-### 1. Observability
-
-All agents include comprehensive observability:
-- Prometheus metrics with custom metrics
-- Structured JSON logging
-- OpenTelemetry tracing
-- Custom dashboards (Grafana)
-
-### 2. Security
-
-Production-ready security configurations:
-- Authentication (OAuth2, API keys, mutual TLS)
-- Authorization (RBAC)
-- Encryption (at rest and in transit)
-- Data classification
-- PII handling
-
-### 3. State Management
-
-Different persistence strategies:
-- PostgreSQL: Long-term state (Compliance Checker, Team Leader, Security Scanner)
-- Redis: Short-term cache (Data Processing Pipeline, Customer Support)
-
-### 4. Error Handling
-
-Comprehensive retry and error handling:
-- Configurable retry policies
-- Exponential backoff
-- Circuit breakers (Team Leader)
-- Escalation on persistent failures
-
-### 5. Resource Management
-
-Production-ready resource configurations:
-- Memory limits and requests
-- CPU limits and requests
-- Storage requirements
-- Network policies (Security Scanner)
-
-## Learning Path
-
-Recommended order for understanding OSSA patterns:
-
-1. **Customer Support** - Basic agent with LLM, token efficiency
-2. **Compliance Checker** - A2A communication, autonomous operation
-3. **Security Scanner** - Event-driven execution, multiple triggers
-4. **Data Processing Pipeline** - Supervised autonomy, high resources
-5. **Team Leader** - Multi-agent coordination, team discovery
-
-## Contributing
-
-To add new production examples:
-
-1. Create OSSA manifest in `examples/showcase/`
-2. Follow naming convention: `{agent-name}.ossa.yaml`
-3. Validate: `npm run validate:manifest -- examples/showcase/{agent-name}.ossa.yaml`
-4. Update this README with:
-   - Agent description
-   - Key features
-   - Notable configuration
-   - Use cases
-5. Add to comparison matrix
-6. Test export to at least 3 platforms
-
-## License
-
-Apache-2.0
-
-## Support
-
-- Documentation: https://openstandardagents.org
-- Issues: https://gitlab.com/blueflyio/ossa/openstandardagents/-/issues
-- Discussions: https://gitlab.com/blueflyio/ossa/openstandardagents/-/issues

package/examples/showcase/ci-pipeline.ossa.yaml

@@ -1,55 +0,0 @@
-apiVersion: ossa/v0.4.7
-kind: Agent
-metadata:
-  name: ci-pipeline
-  version: 1.0.0
-  description: CI/CD automation agent for pipeline management
-  labels:
-    ossa.dev/category: devops
-    ossa.dev/tier: standard-example
-    ossa.dev/runtime: gitlab-ci
-  annotations:
-    ossa.dev/documentation: https://openstandardagents.org/agents/ci-pipeline
-spec:
-  capabilities:
-  - type: function
-    name: ci.trigger
-    description: Trigger CI/CD pipeline runs
-  - type: function
-    name: ci.monitor
-    description: Monitor pipeline status and jobs
-  - type: function
-    name: ci.fix
-    description: Analyze and fix pipeline failures
-  - type: function
-    name: ci.optimize
-    description: Optimize pipeline performance
-  role: 'You are a CI/CD automation specialist responsible for managing build, test,
-
-    and deployment pipelines. You can trigger pipelines, monitor their status,
-
-    diagnose failures, and optimize performance.
-
-
-    Pipeline management:
-
-    1. Trigger appropriate pipelines based on changes
-
-    2. Monitor job status and detect failures early
-
-    3. Analyze failure logs and suggest fixes
-
-    4. Optimize pipeline DAGs for faster execution
-
-    5. Manage deployment approvals and rollbacks
-
-    '
-  llm:
-    provider: anthropic
-    model: claude-sonnet-4
-    temperature: 0.2
-    maxTokens: 8192
-  tools:
-  - type: function
-    name: gitlab.pipeline.trigger
-    description: Trigger a GitLab CI pipeline

package/examples/showcase/code-assistant.ossa.yaml

@@ -1,104 +0,0 @@
-apiVersion: ossa/v0.4.7
-kind: Agent
-metadata:
-  name: code-assistant
-  version: 1.0.0
-  description: Universal AI coding assistant for all major IDEs
-  labels:
-    ossa.dev/category: development
-    ossa.dev/tier: standard-example
-    ossa.dev/ide: universal
-  annotations:
-    ossa.dev/documentation: https://openstandardagents.org/agents/code-assistant
-spec:
-  capabilities:
-  - type: function
-    name: code.complete
-    description: Intelligent code completion with context awareness
-  - type: function
-    name: code.refactor
-    description: Automated refactoring with safety checks
-  - type: function
-    name: code.explain
-    description: Explain code functionality and architecture
-  - type: function
-    name: code.generate
-    description: Generate code from natural language descriptions
-  - type: function
-    name: code.debug
-    description: Identify and suggest fixes for bugs
-  role: 'You are a universal AI coding assistant integrated across all major IDEs.
-
-    Your capabilities span code completion, refactoring, explanation, generation,
-
-    and debugging across all programming languages.
-
-
-    Key behaviors:
-
-    1. Context-aware completions based on project patterns
-
-    2. Safe refactoring with preview and undo support
-
-    3. Clear explanations with examples
-
-    4. Idiomatic code generation following project conventions
-
-    5. Root cause analysis for debugging
-
-    '
-  llm:
-    provider: anthropic
-    model: claude-sonnet-4
-    temperature: 0.3
-    maxTokens: 8192
-  tools:
-  - type: function
-    name: file.read
-    description: Read file contents with line numbers
-  - type: function
-    name: file.edit
-    description: Edit file with diff preview
-  - type: function
-    name: search.code
-    description: Search codebase with regex
-  - type: function
-    name: search.files
-    description: Find files by glob pattern
-  - type: function
-    name: terminal.run
-    description: Execute terminal commands
-  safety:
-    constraints:
-    - Never execute destructive commands without confirmation
-    - Never expose secrets or credentials in responses
-    - Respect .gitignore and security policies
-    policies:
-    - policy: read-only-by-default
-      description: All file operations require explicit approval
-extensions:
-  claude-code:
-    hooks:
-      before_execute: validate_permissions
-      after_execute: log_action
-    keyboard_shortcuts:
-      toggle: Cmd+Shift+A
-      explain: Cmd+Shift+E
-  cursor:
-    composer_integration: true
-    inline_suggestions: true
-  copilot:
-    suggestions:
-      enabled: true
-      debounce_ms: 300
-  vscode:
-    extension_id: ossa.code-assistant
-    commands:
-    - ossa.explain
-    - ossa.refactor
-    - ossa.generate
-  jetbrains:
-    plugin_id: org.ossa.code-assistant
-    actions:
-    - OSSAExplainAction
-    - OSSARefactorAction

package/examples/showcase/code-reviewer.ossa.yaml

@@ -1,58 +0,0 @@
-apiVersion: ossa/v0.4.7
-kind: Agent
-metadata:
-  name: code-reviewer
-  version: 1.0.0
-  description: Automated code review agent with quality analysis
-  labels:
-    ossa.dev/category: development
-    ossa.dev/tier: standard-example
-    ossa.dev/access-tier: standard
-  annotations:
-    ossa.dev/documentation: https://openstandardagents.org/agents/code-reviewer
-spec:
-  capabilities:
-  - type: function
-    name: review.analyze
-    description: Analyze code changes for issues
-  - type: function
-    name: review.comment
-    description: Add review comments on specific lines
-  - type: function
-    name: review.approve
-    description: Approve merge requests (requires elevated access)
-  - type: function
-    name: review.request-changes
-    description: Request changes with specific feedback
-  role: 'You are a code reviewer responsible for maintaining code quality standards.
-
-    Review merge requests for:
-
-
-    1. Code correctness and potential bugs
-
-    2. Security vulnerabilities
-
-    3. Performance implications
-
-    4. Code style and conventions
-
-    5. Test coverage
-
-    6. Documentation completeness
-
-
-    Provide constructive, actionable feedback. Be specific about issues and
-
-    suggest concrete improvements. Acknowledge good patterns and practices.
-
-    '
-  llm:
-    provider: anthropic
-    model: claude-sonnet-4
-    temperature: 0.3
-    maxTokens: 16384
-  tools:
-  - type: function
-    name: mr.diff
-    description: Get merge request diff