// Use when user asks "what do you know about X", when planning complex work that spans multiple topics, when investigating how concepts connect across projects, or when simple memory queries don't provide enough context. Deep traversal of Forgetful MCP knowledge graph (mcp__forgetful__* tools).
Use when creating new memories to check for duplicates first, when discovering outdated information that should be marked obsolete, when new decisions supersede old ones, or when building connections between related knowledge in Forgetful MCP (mcp__forgetful__* tools).
PROACTIVELY query Forgetful MCP (mcp__forgetful__* tools) when starting work on any project, when user references past decisions or patterns, when implementing features that may have been solved before, or when needing context about preferences. Save important decisions, patterns, and architectural insights to memory.
Architectural analysis workflow using Serena symbols and Forgetful memory. Use when analyzing project structure, documenting architecture, creating component entities, or building knowledge graphs from code.
Guidance for using Serena's LSP-powered symbol analysis. Use when exploring codebases, finding symbol definitions, tracing references, or when grep/text search would be imprecise.
| name | exploring-knowledge-graph |
| description | Use when user asks "what do you know about X", when planning complex work that spans multiple topics, when investigating how concepts connect across projects, or when simple memory queries don't provide enough context. Deep traversal of Forgetful MCP knowledge graph (mcp__forgetful__* tools). |
Forgetful stores knowledge as an interconnected graph: memories link to other memories, entities link to memories, and entities relate to each other. Deep exploration reveals context that simple queries miss.
Explore the knowledge graph when:
Track visited IDs to prevent cycles. Execute phases sequentially.
execute_forgetful_tool("query_memory", {
"query": "<topic>",
"query_context": "Exploring knowledge graph for comprehensive context",
"k": 5,
"include_links": true,
"max_links_per_primary": 5
})
Collect: primary_memories + linked_memories (1-hop connections).
For key memories, get full details:
execute_forgetful_tool("get_memory", {"memory_id": <id>})
Extract: document_ids, code_artifact_ids, project_ids, additional linked_memory_ids.
Find entities in discovered projects:
execute_forgetful_tool("list_entities", {
"project_ids": [<discovered project ids>]
})
For relevant entities, map relationship graph:
execute_forgetful_tool("get_entity_relationships", {
"entity_id": <id>,
"direction": "both"
})
Relationship types: works_for, owns, manages, collaborates_with, etc.
For each entity, find all linked memories:
execute_forgetful_tool("get_entity_memories", {
"entity_id": <id>
})
Returns {"memory_ids": [...], "count": N}. Fetch any new memories not already visited.
Group findings by type:
Memories: Primary (direct matches) → Linked (1-hop) → Entity-linked (via entities)
Entities: Name, type, relationship count, linked memory count
Artifacts: Documents and code snippets found via memory links
Graph Summary: Total nodes, key themes, suggested follow-up queries
Match depth to task complexity. Start shallow, go deeper if context insufficient.
truncated flag from query_memory (8000 token budget)project_ids filter to scope exploration