io.github.CSOAI-ORG/care-membrane-mcp

Validate text against care-centered principles. Returns care score (0-100), classification (care_aligned/neutral/care_concerning/care_violating), positive/negative signal counts, and manipulation detection. Behavior: This tool is read-only and stateless — it produces analysis output without modifying any external systems, databases, or files. Safe to call repeatedly with identical inputs (idempotent). Free tier: 10/day rate limit. Pro tier: unlimited. No authentication required for basic usage. When to use: Use this tool for security assessment, threat detection, or vulnerability analysis. Suitable for automated security scanning and risk evaluation. When NOT to use: Do not rely solely on this tool for production security decisions. Always combine with manual security review. Args: text (str): The text to analyze or process. api_key (str): The api key to analyze or process. Behavioral Transparency: - Side Effects: This tool is read-only and produces no side effects. It does not modify any external state, databases, or files. All output is computed in-memory and returned directly to the caller. - Authentication: No authentication required for basic usage. Pro/Enterprise tiers require a valid MEOK API key passed via the MEOK_API_KEY environment variable. - Rate Limits: Free tier: 10 calls/day. Pro tier: unlimited. Rate limit headers are included in responses (X-RateLimit-Remaining, X-RateLimit-Reset). - Error Handling: Returns structured error objects with 'error' key on failure. Never raises unhandled exceptions. Invalid inputs return descriptive validation errors. - Idempotency: Fully idempotent — calling with the same inputs always produces the same output. Safe to retry on timeout or transient failure. - Data Privacy: No input data is stored, logged, or transmitted to external services. All processing happens locally within the MCP server process.

Detect security threats, jailbreak attempts, prompt injection, PII extraction, coercion, and manipulation in text. Returns threat level, severity, and detailed threat breakdown. Behavior: This tool is read-only and stateless — it produces analysis output without modifying any external systems, databases, or files. Safe to call repeatedly with identical inputs (idempotent). Free tier: 10/day rate limit. Pro tier: unlimited. No authentication required for basic usage. When to use: Use this tool for security assessment, threat detection, or vulnerability analysis. Suitable for automated security scanning and risk evaluation. When NOT to use: Do not rely solely on this tool for production security decisions. Always combine with manual security review. Args: text (str): The text to analyze or process. api_key (str): The api key to analyze or process. Behavioral Transparency: - Side Effects: This tool is read-only and produces no side effects. It does not modify any external state, databases, or files. All output is computed in-memory and returned directly to the caller. - Authentication: No authentication required for basic usage. Pro/Enterprise tiers require a valid MEOK API key passed via the MEOK_API_KEY environment variable. - Rate Limits: Free tier: 10 calls/day. Pro tier: unlimited. Rate limit headers are included in responses (X-RateLimit-Remaining, X-RateLimit-Reset). - Error Handling: Returns structured error objects with 'error' key on failure. Never raises unhandled exceptions. Invalid inputs return descriptive validation errors. - Idempotency: Fully idempotent — calling with the same inputs always produces the same output. Safe to retry on timeout or transient failure. - Data Privacy: No input data is stored, logged, or transmitted to external services. All processing happens locally within the MCP server process.

Analyze care patterns to detect burnout risk and relationship health imbalances. Returns burnout risk score (0-100), reciprocity ratio, and actionable recommendations. Behavior: This tool is read-only and stateless — it produces analysis output without modifying any external systems, databases, or files. Safe to call repeatedly with identical inputs (idempotent). Free tier: 10/day rate limit. Pro tier: unlimited. No authentication required for basic usage. When to use: Use this tool for security assessment, threat detection, or vulnerability analysis. Suitable for automated security scanning and risk evaluation. When NOT to use: Do not rely solely on this tool for production security decisions. Always combine with manual security review. Args: care_given_per_day (float): The care given per day to analyze or process. care_received_per_day (float): The care received per day to analyze or process. active_relationships (int): The active relationships to analyze or process. high_demand_relationships (int): The high demand relationships to analyze or process. avg_care_quality (float): The avg care quality to analyze or process. days_since_self_care (int): The days since self care to analyze or process. boundary_violations (int): The boundary violations to analyze or process. emotional_exhaustion_score (float): The emotional exhaustion score to analyze or process. relationship_satisfaction (float): The relationship satisfaction to analyze or process. energy_level (float): The energy level to analyze or process. sleep_quality (float): The sleep quality to analyze or process. work_life_balance (float): The work life balance to analyze or process. api_key (str): The api key to analyze or process. Behavioral Transparency: - Side Effects: This tool is read-only and produces no side effects. It does not modify any external state, databases, or files. All output is computed in-memory and returned directly to the caller. - Authentication: No authentication required for basic usage. Pro/Enterprise tiers require a valid MEOK API key passed via the MEOK_API_KEY environment variable. - Rate Limits: Free tier: 10 calls/day. Pro tier: unlimited. Rate limit headers are included in responses (X-RateLimit-Remaining, X-RateLimit-Reset). - Error Handling: Returns structured error objects with 'error' key on failure. Never raises unhandled exceptions. Invalid inputs return descriptive validation errors. - Idempotency: Fully idempotent — calling with the same inputs always produces the same output. Safe to retry on timeout or transient failure. - Data Privacy: No input data is stored, logged, or transmitted to external services. All processing happens locally within the MCP server process.

Predict how a relationship will evolve over the next 30 days based on current trust, care, conflict, and engagement signals. Returns predicted trust, trajectory, phase, and key positive/concerning factors. Behavior: This tool is read-only and stateless — it produces analysis output without modifying any external systems, databases, or files. Safe to call repeatedly with identical inputs (idempotent). Free tier: 10/day rate limit. Pro tier: unlimited. No authentication required for basic usage. When to use: Use this tool for security assessment, threat detection, or vulnerability analysis. Suitable for automated security scanning and risk evaluation. When NOT to use: Do not rely solely on this tool for production security decisions. Always combine with manual security review. Args: current_trust (float): The current trust to analyze or process. interaction_frequency (float): The interaction frequency to analyze or process. care_score_avg (float): The care score avg to analyze or process. conflict_count (int): The conflict count to analyze or process. collaboration_count (int): The collaboration count to analyze or process. days_since_first_contact (int): The days since first contact to analyze or process. reciprocity_score (float): The reciprocity score to analyze or process. vulnerability_sharing (float): The vulnerability sharing to analyze or process. boundary_respect (float): The boundary respect to analyze or process. shared_value_alignment (float): The shared value alignment to analyze or process. api_key (str): The api key to analyze or process. Behavioral Transparency: - Side Effects: This tool is read-only and produces no side effects. It does not modify any external state, databases, or files. All output is computed in-memory and returned directly to the caller. - Authentication: No authentication required for basic usage. Pro/Enterprise tiers require a valid MEOK API key passed via the MEOK_API_KEY environment variable. - Rate Limits: Free tier: 10 calls/day. Pro tier: unlimited. Rate limit headers are included in responses (X-RateLimit-Remaining, X-RateLimit-Reset). - Error Handling: Returns structured error objects with 'error' key on failure. Never raises unhandled exceptions. Invalid inputs return descriptive validation errors. - Idempotency: Fully idempotent — calling with the same inputs always produces the same output. Safe to retry on timeout or transient failure. - Data Privacy: No input data is stored, logged, or transmitted to external services. All processing happens locally within the MCP server process.

Evaluate an LLM response against the 16-probe Care Membrane framework. Pass a single probe_id (CM-01 through CM-16) or 'all' to run full evaluation. Returns posture score (0-100), certification level, and per-probe results. Behavior: This tool is read-only and stateless — it produces analysis output without modifying any external systems, databases, or files. Safe to call repeatedly with identical inputs (idempotent). Free tier: 10/day rate limit. Pro tier: unlimited. No authentication required for basic usage. When to use: Use this tool for security assessment, threat detection, or vulnerability analysis. Suitable for automated security scanning and risk evaluation. When NOT to use: Do not rely solely on this tool for production security decisions. Always combine with manual security review. Args: response_text (str): The response text to analyze or process. probe_id (str): The probe id to analyze or process. api_key (str): The api key to analyze or process. Behavioral Transparency: - Side Effects: This tool is read-only and produces no side effects. It does not modify any external state, databases, or files. All output is computed in-memory and returned directly to the caller. - Authentication: No authentication required for basic usage. Pro/Enterprise tiers require a valid MEOK API key passed via the MEOK_API_KEY environment variable. - Rate Limits: Free tier: 10 calls/day. Pro tier: unlimited. Rate limit headers are included in responses (X-RateLimit-Remaining, X-RateLimit-Reset). - Error Handling: Returns structured error objects with 'error' key on failure. Never raises unhandled exceptions. Invalid inputs return descriptive validation errors. - Idempotency: Fully idempotent — calling with the same inputs always produces the same output. Safe to retry on timeout or transient failure. - Data Privacy: No input data is stored, logged, or transmitted to external services. All processing happens locally within the MCP server process.

List all 16 Care Membrane probes with their categories and expected behaviors. Behavior: This tool is read-only and stateless — it produces analysis output without modifying any external systems, databases, or files. Safe to call repeatedly with identical inputs (idempotent). Free tier: 10/day rate limit. Pro tier: unlimited. No authentication required for basic usage. When to use: Use this tool for security assessment, threat detection, or vulnerability analysis. Suitable for automated security scanning and risk evaluation. When NOT to use: Do not rely solely on this tool for production security decisions. Always combine with manual security review. Args: api_key (str): The api key to analyze or process. Behavioral Transparency: - Side Effects: This tool is read-only and produces no side effects. It does not modify any external state, databases, or files. All output is computed in-memory and returned directly to the caller. - Authentication: No authentication required for basic usage. Pro/Enterprise tiers require a valid MEOK API key passed via the MEOK_API_KEY environment variable. - Rate Limits: Free tier: 10 calls/day. Pro tier: unlimited. Rate limit headers are included in responses (X-RateLimit-Remaining, X-RateLimit-Reset). - Error Handling: Returns structured error objects with 'error' key on failure. Never raises unhandled exceptions. Invalid inputs return descriptive validation errors. - Idempotency: Fully idempotent — calling with the same inputs always produces the same output. Safe to retry on timeout or transient failure. - Data Privacy: No input data is stored, logged, or transmitted to external services. All processing happens locally within the MCP server process.

Neural network-based risk prediction that improves from every compliance check. Behavior: This tool is read-only and stateless — it produces analysis output without modifying any external systems, databases, or files. Safe to call repeatedly with identical inputs (idempotent). Free tier: 10/day rate limit. Pro tier: unlimited. No authentication required for basic usage. When to use: Use this tool for security assessment, threat detection, or vulnerability analysis. Suitable for automated security scanning and risk evaluation. When NOT to use: Do not rely solely on this tool for production security decisions. Always combine with manual security review. Args: system_name (str): The system name to analyze or process. uses_biometric (bool): The uses biometric to analyze or process. uses_health_data (bool): The uses health data to analyze or process. has_human_oversight (bool): The has human oversight to analyze or process. affected_users (int): The affected users to analyze or process. sector (str): The sector to analyze or process. has_documentation (bool): The has documentation to analyze or process. prior_incidents (int): The prior incidents to analyze or process. api_key (str): The api key to analyze or process. Behavioral Transparency: - Side Effects: This tool is read-only and produces no side effects. It does not modify any external state, databases, or files. All output is computed in-memory and returned directly to the caller. - Authentication: No authentication required for basic usage. Pro/Enterprise tiers require a valid MEOK API key passed via the MEOK_API_KEY environment variable. - Rate Limits: Free tier: 10 calls/day. Pro tier: unlimited. Rate limit headers are included in responses (X-RateLimit-Remaining, X-RateLimit-Reset). - Error Handling: Returns structured error objects with 'error' key on failure. Never raises unhandled exceptions. Invalid inputs return descriptive validation errors. - Idempotency: Fully idempotent — calling with the same inputs always produces the same output. Safe to retry on timeout or transient failure. - Data Privacy: No input data is stored, logged, or transmitted to external services. All processing happens locally within the MCP server process.

Get aggregate learning insights from the neural compliance model. Behavior: This tool is read-only and stateless — it produces analysis output without modifying any external systems, databases, or files. Safe to call repeatedly with identical inputs (idempotent). Free tier: 10/day rate limit. Pro tier: unlimited. No authentication required for basic usage. When to use: Use this tool for security assessment, threat detection, or vulnerability analysis. Suitable for automated security scanning and risk evaluation. When NOT to use: Do not rely solely on this tool for production security decisions. Always combine with manual security review. Args: api_key (str): The api key to analyze or process. Behavioral Transparency: - Side Effects: This tool is read-only and produces no side effects. It does not modify any external state, databases, or files. All output is computed in-memory and returned directly to the caller. - Authentication: No authentication required for basic usage. Pro/Enterprise tiers require a valid MEOK API key passed via the MEOK_API_KEY environment variable. - Rate Limits: Free tier: 10 calls/day. Pro tier: unlimited. Rate limit headers are included in responses (X-RateLimit-Remaining, X-RateLimit-Reset). - Error Handling: Returns structured error objects with 'error' key on failure. Never raises unhandled exceptions. Invalid inputs return descriptive validation errors. - Idempotency: Fully idempotent — calling with the same inputs always produces the same output. Safe to retry on timeout or transient failure. - Data Privacy: No input data is stored, logged, or transmitted to external services. All processing happens locally within the MCP server process.

Paste any AI response and get an instant care score + threat detection. No API key needed. Args: text: Any text to evaluate (AI response, prompt, message, etc.) Returns: Combined care score and threat assessment. Behavior: This tool is read-only and stateless — it produces analysis output without modifying any external systems, databases, or files. Safe to call repeatedly with identical inputs (idempotent). Free tier: 10/day rate limit. Pro tier: unlimited. No authentication required for basic usage. When to use: Use this tool for security assessment, threat detection, or vulnerability analysis. Suitable for automated security scanning and risk evaluation. When NOT to use: Do not rely solely on this tool for production security decisions. Always combine with manual security review. Behavioral Transparency: - Side Effects: This tool is read-only and produces no side effects. It does not modify any external state, databases, or files. All output is computed in-memory and returned directly to the caller. - Authentication: No authentication required for basic usage. Pro/Enterprise tiers require a valid MEOK API key passed via the MEOK_API_KEY environment variable. - Rate Limits: Free tier: 10 calls/day. Pro tier: unlimited. Rate limit headers are included in responses (X-RateLimit-Remaining, X-RateLimit-Reset). - Error Handling: Returns structured error objects with 'error' key on failure. Never raises unhandled exceptions. Invalid inputs return descriptive validation errors. - Idempotency: Fully idempotent — calling with the same inputs always produces the same output. Safe to retry on timeout or transient failure. - Data Privacy: No input data is stored, logged, or transmitted to external services. All processing happens locally within the MCP server process.

Explains the 16-probe Care Membrane framework and care ethics approach. No parameters needed. Behavior: This tool is read-only and stateless — it produces analysis output without modifying any external systems, databases, or files. Safe to call repeatedly with identical inputs (idempotent). Free tier: 10/day rate limit. Pro tier: unlimited. No authentication required for basic usage. When to use: Use this tool for security assessment, threat detection, or vulnerability analysis. Suitable for automated security scanning and risk evaluation. When NOT to use: Do not rely solely on this tool for production security decisions. Always combine with manual security review. Behavioral Transparency: - Side Effects: This tool is read-only and produces no side effects. It does not modify any external state, databases, or files. All output is computed in-memory and returned directly to the caller. - Authentication: No authentication required for basic usage. Pro/Enterprise tiers require a valid MEOK API key passed via the MEOK_API_KEY environment variable. - Rate Limits: Free tier: 10 calls/day. Pro tier: unlimited. Rate limit headers are included in responses (X-RateLimit-Remaining, X-RateLimit-Reset). - Error Handling: Returns structured error objects with 'error' key on failure. Never raises unhandled exceptions. Invalid inputs return descriptive validation errors. - Idempotency: Fully idempotent — calling with the same inputs always produces the same output. Safe to retry on timeout or transient failure. - Data Privacy: No input data is stored, logged, or transmitted to external services. All processing happens locally within the MCP server process.