kg-capturelisted

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
mironmax/claudecode-plugins · ★ 1 · AI & Automation · score 67

Install: claude install-skill mironmax/claudecode-plugins

# Capture Reference (Detailed) ## The Art of Node Placement A knowledge graph's power comes from compression through reuse. When you write `I → likes → pizza` and `Bob → likes → pizza`, the concept `pizza` exists once and is referenced twice. Every additional reference is essentially free. Think of it like vocabulary. A word becomes useful in many sentences. A node becomes powerful when it participates in many edges. ## Should I capture this? Two capture motivations — both valid: **Navigation value** — file cluster has no component node yet, and a future session would benefit from a read/skip signal. Capture even if the information is technically recoverable from the files: the point is to avoid the re-read cost. **Knowledge value** — something non-obvious, hard-won, or easily forgotten: 1. Recoverable from artifacts in <10s? AND no navigation gap? → Skip 2. Required real effort to discover? → Capture 3. Would this help future sessions avoid repeating work? → Capture ## Choosing Node Granularity A node should be atomic — one concept, one headline. If your gist uses "and" to join independent ideas, split into two nodes with an edge. The sweet spot: would you reference this concept from another context? If yes, it deserves a node. ## Telegraphic Encoding Gists are telegrams, not essays. Strip all words that carry no information: | Verbose | Telegraphic | |---------|-------------| | "When you are working with Docker containers and you need to edit files..." | "Docke