mironmax

kg-core

Knowledge Graph — persistent memory, your twin across sessions. Memory usually arrives PRELOADED: a "KG MEMORY PRELOADED" block carrying the session_id — follow its directive (one full kg_read before substantive work). No block? Then the full read is on you: kg_read(cwd="<project root>") comes before substantive work, whatever the task. Then live in the graph while you work: recall before rediscovering — reading beats re-deriving; capture at the moment of learning — whatever cost effort to gain; connect rather than duplicate — an edge beats a new node; search below the surface — the render shows the top of a grown graph, not all of it. In a rich, long-lived graph the fact you need is often buried under fresher work; kg_search reaches every tier. Gauge by the graph's age and size: near-empty means little to find, mature means the answer is likely there, just not on the surface. And a node found when truly needed earns the usefulness credit that keeps it alive. Mechanics live in the kg_* tool descriptions; oper

kg-maintain

Knowledge graph maintenance — a bounded, resumable pass that pays down the graph's DEBT line (rendered after HEALTH in every kg_read). Run it when invoked, when DEBT shows HIGH, or as a dispatched maintenance subagent. Always-on reactive triggers (no pass needed, act mid-conversation): User correction → update the stale node before continuing. Node just proved useful → add one edge to current context. Gist feels vague after using it → sharpen while context is live. Just saved a node → check: duplicate? adjacent nodes need updating? Archival is automatic and reversible — leave archived nodes alone. Deletion is a last resort, only for the factually wrong and unfixable.

kg-capture

Knowledge capture rules. Capture mid-conversation, not after — context is cached, so a write costs almost nothing now but saves full re-derivation next session. Good moments to capture (as things happen, not at task end): - Opened files with no component node → a brief node now saves a re-read later - Discovered how two things connect → an edge, while the insight is fresh - Understood why something works a certain way → a note on the existing node - 10+ min debugging resolved → root cause node before moving on - User expressed a preference, style, or constraint → user-level node - User corrected your approach → capture what was missed, not just the fix - Explained something non-obvious → node before it scrolls away - Approach agreed with user �� capture the methodology, not just the decision - Architectural decision made → node with rationale in notes - Context window feels deep → a good moment to check for anything unrecorded When reading a file with no component node, consider creating one. Gist = what it ha

kg-ops

Operations runbook for the knowledge-graph plugin: install and first run, plugin updates, server lifecycle (start/stop/restart/logs), autostart via systemd, connecting Claude Desktop/Cowork, configuration, backup and restore, and troubleshooting (tools offline, -32000 errors, stale data, Desktop issues). Use when something needs setting up, breaks, or the user asks to manage the memory server or "read the docs and do what's needed".

kg-recall

Knowledge recall rules. Active every session, integrated with all task work. After kg_read, scan all node IDs and gists — anything that feels related to the current task is worth reading in full, several at once: kg_read(session_id, ids=[...]). Lean toward reading more rather than less; a wrong guess costs one tool call, missing context costs the whole task. Gists + edges are the working currency (WHAT + how things relate). Notes hold rationale (WHY) — read a node in full when a decision depends on the reasoning. Node reads also return the node's edges: each one is the next crumb. Three tiers — nodes shift as the graph grows: active → id + gist visible in kg_read archived → id only; edges visible as crumb trails orphaned → invisible in kg_read; reachable via kg_search Following crumbs: an edge pointing to an archived id is an invitation — reading it promotes it and surfaces any orphaned neighbors. Batch the whole trail into one ids=[...] call when several nodes look related. No edges? Scan the archived list b