Axom MCP Server

Store, retrieve, search, and manage persistent memories in the Axom database.

Memory Types:

• long_term: Reusable patterns, architectural decisions, gotchas

• short_term: Task-specific context, debug notes, current task state

• reflex: Learned heuristics ("Always check X before Y" patterns)

• dreams: Experimental ideas, creative explorations

Naming Convention: [type][descriptor][YYYYMMDD] Example: bugfix_auth_timeout_20260203

Content Format (recommended): TASK|APPROACH|OUTCOME|GOTCHAS|RELATED

Actions:

• write: Store a new memory

• read: Retrieve a specific memory by name

• list: List memories with optional filters

• search: Full-text search across memories

• delete: Remove a memory by name

Execute file operations and shell commands with chain-reaction support.

Operations:

• read: Read file contents from allowed directories

• write: Write data to files (unless AXOM_READ_ONLY=true)

• shell: Execute shell commands (unless AXOM_READ_ONLY=true)

Chain Reactions: Chain multiple operations together using the chain parameter. Each step can reference the previous result using ${_result} variable substitution.

Example: { "operation": "read", "target": "/file.txt", "chain": [ { "tool": "axom_mcp_transform", "args": {"input": "${_result.content}", "output_format": "json"} } ] }

Security:

• File operations restricted to allowed directories (cwd, ~/)

• Shell/write operations enabled by default (set AXOM_READ_ONLY=true to disable)

• Input size limits: 10MB max for files

Analyze code and data with configurable depth and scope.

Analysis Types:

• debug: Troubleshoot issues, investigate errors, diagnose problems

• review: Code review, quality assessment, best practices

• audit: Security audit, compliance check, vulnerability scan

• refactor: Refactoring suggestions, code improvement recommendations

• test: Test coverage analysis, test generation suggestions

Focus Areas:

• security: Security vulnerabilities, injection risks, auth issues

• performance: Performance bottlenecks, optimization opportunities

• architecture: Architectural patterns, design issues

• maintainability: Code smell, complexity, documentation

Depth Levels:

• minimal: Quick scan, critical issues only

• low: Basic analysis, obvious issues

• medium: Standard analysis (default)

• high: Deep analysis, all issues

• max: Exhaustive analysis, edge cases

Chain Support: Use chain parameter to automatically act on analysis results.

Discover available resources, structures, and capabilities.

Discovery Domains:

• files: List and search files in allowed directories

• tools: List available MCP tools and their capabilities

• memory: Explore memory structure and statistics

• capabilities: Check server capabilities and configuration

• all: Comprehensive discovery across all domains

Filter Options:

• pattern: Glob pattern for file filtering (e.g., *.py)

• type: File type filter (file, directory, all)

• memory_type: Filter memories by type

• importance: Filter memories by importance

Chain Support: Use chain parameter to act on discovered resources.

Transform data between formats and structures.

Supported Formats:

• json: JSON objects and arrays

• yaml: YAML documents

• csv: Comma-separated values

• markdown: Markdown documents

• code: Source code (with language detection)

Transformation Rules:

• field_mapping: Rename or restructure fields

• filter: Include/exclude specific fields

• sort: Sort arrays by field

• aggregate: Group and aggregate data

Template Support: Use Jinja2 templates for custom output formatting.

Chain Support: Use chain parameter to continue processing transformed data.