LifeOS MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| PYTHONUNBUFFERED | No | Set to 1 to disable Python output buffering. | 1 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| lifeos_get_contextA | Return a compact, deterministic bootstrap context snapshot for the user. Reads core canonical files directly (profile, communication instructions, current goals, and active project overviews). This is a deterministic bootstrap operation that does not perform vector search or dump the vault. |
| lifeos_searchA | Perform semantic vector search across indexed vault files. Returns relevant text passages with explicit provenance: [Source: | Heading: | Similarity: ]. Args: query: Natural language query to search for. category: Optional vault category filter (e.g. 'projects', 'goals', 'health', 'learning'). top_k: Number of relevant chunks to return (default: 5). |
| lifeos_read_fileA | Read the exact contents of a specific Markdown file from the vault. Strict path validation prevents directory traversal outside the vault root. Args: relative_path: Vault-relative path (e.g. 'projects/chronolog.md', 'identity/values.md'). |
| lifeos_get_instructionsA | Retrieve the user's explicit personal instructions from identity/how_i_want_ai_to_treat_me.md. |
| lifeos_list_topicsA | List all vault categories and tracked Markdown files in a concise hierarchy. |
| lifeos_add_memoryA | Store an agent-generated memory entry under vault/memory/. Restricted write operation:
Args: category: Semantic category metadata tag (e.g. 'work', 'preferences', 'decisions'). filename: Kebab-case filename ending with .md (e.g. '2026-10-02-backend-priority.md'). content: The Markdown text to record. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 6 tools
Most tools target clearly distinct operations (semantic search vs exact read vs listing vs write). However, lifeos_get_context and lifeos_get_instructions overlap: get_context already pulls 'communication instructions' while get_instructions pulls 'personal instructions' from a specific identity file, which could cause misselection between the two.
All tools use a consistent snake_case pattern with the lifeos_ prefix (lifeos_get_context, lifeos_search, lifeos_read_file, lifeos_get_instructions, lifeos_list_topics, lifeos_add_memory). Verb-first style is predictable throughout, with only the bare 'search' as a mild deviation.
Six tools is a reasonable, well-scoped set for a personal vault assistant covering bootstrap, discovery, search, read, and memory write. It leans slightly thin, but each tool earns its place without redundancy.
The read/lookup side is well covered (context, search, read, list, instructions), but write operations are restricted to adding memories only. There is no way to update, edit, or delete memory entries or canonical files, leaving notable lifecycle gaps for an agent managing the vault.