القائمة
Query an Emergent project knowledge base using natural language or direct hybrid search. Use when the user wants to search, explore, or ask questions about content in an Emergent project.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Manage Emergent runtime agents and agent definitions — create, trigger, monitor runs, respond to questions, and manage agent configurations. Use when the user wants to run, inspect, or configure AI agents in an Emergent project.
Apply and export Emergent Blueprints — declarative directories of template packs, agent definitions, and seed data applied with `memory blueprints`. Use when setting up a project from a blueprint, seeding graph data, or exporting an existing project's graph as re-applyable seed files.
Full Memory CLI command reference with all subcommands and flags. Use when you need exact command syntax, flag names, or usage examples for any `memory` CLI command.
Manage MCP (Model Context Protocol) servers in an Emergent project — register, inspect, sync tools, and delete servers. Use when the user wants to add, configure, or troubleshoot MCP servers connected to their Emergent project.
Onboard a project into Memory — understand what the project is, choose or create a Memory project, design and install a schema (template pack), then guide on creating objects and relationships. Use when setting up Memory for a new project or codebase for the first time.
Manage LLM provider credentials and browse available models in Emergent. Use when the user wants to configure API keys, set up Vertex AI, check which models are available, or review LLM usage and costs.
| name | memory-query |
| description | Query an Emergent project knowledge base using natural language or direct hybrid search. Use when the user wants to search, explore, or ask questions about content in an Emergent project. |
| metadata | {"author":"emergent","version":"2.0"} |
New to Emergent? Load the
memory-onboardskill first to set up the project's knowledge graph schema before querying.
Query an Emergent project knowledge base using the memory query command.
memory browse — it launches a full interactive TUI that blocks on terminal input and will hang in an automated agent context.memory commands with NO_PROMPT=1 (e.g. NO_PROMPT=1 memory <cmd>). Without it, the CLI may show interactive pickers when no project, agent, MCP server, skill, or agent-definition ID is provided. Do not add this to .env.local — it must only apply to agent-driven invocations.--project <id> on project-scoped commands, or ensure MEMORY_PROJECT is set.Agent mode (default) — AI reasoning over the knowledge graph. Best for complex or multi-hop questions.
memory query "who directed fight club and what are their other movies?"
memory query --project <id> "list all requirements for the auth module"
memory query --show-tools "what are the key relationships between X and Y?"
Search mode — Direct hybrid search (semantic + lexical), no AI reasoning. Best for finding specific content fast.
memory query --mode=search "fight club"
memory query --mode=search --result-types=graph --limit=20 "authentication"
memory query --mode=search --result-types=text "API rate limiting"
| Flag | Default | Notes |
|---|---|---|
--mode | agent | agent or search |
--project | from config | Override target project |
--json | false | Machine-readable output |
--show-tools | false | Show agent tool calls (agent mode only) |
--limit | 10 | Max results (search mode only) |
--result-types | both | graph, text, or both (search mode only) |
--fusion-strategy | weighted | weighted, rrf, interleave, graph_first, text_first |
--project is set, the CLI uses the default project from config; ask the user to confirm or specify one if the context is ambiguous--output json + --json for downstream processing or to pass results to another toolAgent mode streams the response and prints a final answer. Search mode returns a table (or JSON) of matching objects/documents with scores.
Use --json to get structured output suitable for parsing:
memory query --mode=search --json "authentication" | jq '.[].title'