Knowledge Graph

**Quick judgement** A focused, file-based knowledge-graph maintenance tool that is well-suited for OpenClaw/Clawdbot setups using the `life/areas/**` structure. It offers deterministic, script-driven updates (add, supersede, summarize) and automatic ID management, making it significantly safer and more reliable than manual JSON editing for this specific use case. **Strengths** - Strong fit for maintaining long-term, structured memory in `life/areas/<kind>/<slug>/` directories. - Append-only with superseding, preserving history while clearly marking obsolete facts. - Auto-incremented, per-entity IDs reduce the risk of collisions and inconsistent references. - Summary regeneration from active facts keeps `summary.md` aligned with the underlying items. **Risks / Limitations** - Tightly coupled to the expected directory and file layout; not a general-purpose KG solution. - Incorrect entity paths, categories, or sloppy fact phrasing can still introduce semantic errors into the knowledge base. - Superseding requires specifying the correct old ID; mistakes there can lead to confusing history or unsuperseded stale facts. - Large `items.json` files might eventually impact performance or merge ergonomics in multi-user setups, though the design is simple enough to manage. **Recommended scenarios** - You are building or extending Clawdbot and need a deterministic way for agents to update persistent knowledge without hand-editing JSON. - You want an append-only, auditable record of changes where facts are superseded, not deleted. - You need to periodically regenerate consistent human-readable summaries of entities from their active facts. - You prefer simple, file-based infrastructure (Git-friendly) over a database-backed knowledge graph for your current scale.