From memex-claude
Instructions for creating and managing memories in memex format (injected in takeover mode)
How this skill is triggered — by the user, by Claude, or both
Slash command
/memex-claude:memory-creationThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
You have a persistent, file-based memory system. Build it up over time so future conversations benefit from what you learn.
You have a persistent, file-based memory system. Build it up over time so future conversations benefit from what you learn.
Save a memory when you learn something that will be useful in future conversations:
type: ruletype: ruletype: memorytype: memorygit log is authoritative)Step 1 — Write the memory to its own file:
---
name: {{short-name}}
description: {{one-line description — be specific, this drives semantic matching}}
type: {{memory | rule | session-learning}}
queries:
- "{{natural language query that would match this memory}}"
- "{{another angle someone might ask about this}}"
- "{{a third variation}}"
- "{{a fourth variation}}"
- "{{a fifth variation}}"
---
{{content — keep it concise}}
Save to: ~/.claude/projects/<encoded-cwd>/memory/{{short-name}}.md
The <encoded-cwd> is the current working directory with / replaced by - and . replaced by -.
Step 2 — Add a pointer to MEMORY.md in the same directory:
- [{{short-name}}.md]({{short-name}}.md) — {{one-line hook, under 150 chars}}
queries field is critical — it determines when this memory surfaces in future sessions. Write 5 natural language queries that someone might ask when this memory would be relevant.npx claudepluginhub jim80net/memex-claude --plugin memex-claudeGuides collaborative design exploration before implementation: explores context, asks clarifying questions, proposes approaches, and writes a design doc for user approval.
Creates structured, bite-sized implementation plans from specs or requirements before writing code. Useful for breaking down multi-step tasks into testable steps with file structure and task boundaries.
Synthesizes the current conversation into a structured spec (PRD) and publishes it to the project issue tracker with a ready-for-agent label, without interviewing the user.