{"skill":{"slug":"memory-setup-1-0-0","displayName":"Memory Setup 1.0.0","summary":"Enable and configure Moltbot/Clawdbot memory search for persistent context. Use when setting up memory, fixing \"goldfish brain,\" or helping users configure m...","description":"---\nname: memory-setup\ndescription: Enable and configure Moltbot/Clawdbot memory search for persistent context. Use when setting up memory, fixing \"goldfish brain,\" or helping users configure memorySearch in their config. Covers MEMORY.md, daily logs, and vector search setup.\n---\n\n# Memory Setup Skill\n\nTransform your agent from goldfish to elephant. This skill helps configure persistent memory for Moltbot/Clawdbot.\n\n## Quick Setup\n\n### 1. Enable Memory Search in Config\n\nAdd to `~/.clawdbot/clawdbot.json` (or `moltbot.json`):\n\n```json\n{\n  \"memorySearch\": {\n    \"enabled\": true,\n    \"provider\": \"voyage\",\n    \"sources\": [\"memory\", \"sessions\"],\n    \"indexMode\": \"hot\",\n    \"minScore\": 0.3,\n    \"maxResults\": 20\n  }\n}\n```\n\n### 2. Create Memory Structure\n\nIn your workspace, create:\n\n```\nworkspace/\n├── MEMORY.md              # Long-term curated memory\n└── memory/\n    ├── logs/              # Daily logs (YYYY-MM-DD.md)\n    ├── projects/          # Project-specific context\n    ├── groups/            # Group chat context\n    └── system/            # Preferences, setup notes\n```\n\n### 3. Initialize MEMORY.md\n\nCreate `MEMORY.md` in workspace root:\n\n```markdown\n# MEMORY.md — Long-Term Memory\n\n## About [User Name]\n- Key facts, preferences, context\n\n## Active Projects\n- Project summaries and status\n\n## Decisions & Lessons\n- Important choices made\n- Lessons learned\n\n## Preferences\n- Communication style\n- Tools and workflows\n```\n\n## Config Options Explained\n\n| Setting | Purpose | Recommended |\n|---------|---------|-------------|\n| `enabled` | Turn on memory search | `true` |\n| `provider` | Embedding provider | `\"voyage\"` |\n| `sources` | What to index | `[\"memory\", \"sessions\"]` |\n| `indexMode` | When to index | `\"hot\"` (real-time) |\n| `minScore` | Relevance threshold | `0.3` (lower = more results) |\n| `maxResults` | Max snippets returned | `20` |\n\n### Provider Options\n- `voyage` — Voyage AI embeddings (recommended)\n- `openai` — OpenAI embeddings\n- `local` — Local embeddings (no API needed)\n\n### Source Options\n- `memory` — MEMORY.md + memory/*.md files\n- `sessions` — Past conversation transcripts\n- `both` — Full context (recommended)\n\n## Daily Log Format\n\nCreate `memory/logs/YYYY-MM-DD.md` daily:\n\n```markdown\n# YYYY-MM-DD — Daily Log\n\n## [Time] — [Event/Task]\n- What happened\n- Decisions made\n- Follow-ups needed\n\n## [Time] — [Another Event]\n- Details\n```\n\n## Agent Instructions (AGENTS.md)\n\nAdd to your AGENTS.md for agent behavior:\n\n```markdown\n## Memory Recall\nBefore answering questions about prior work, decisions, dates, people, preferences, or todos:\n1. Run memory_search with relevant query\n2. Use memory_get to pull specific lines if needed\n3. If low confidence after search, say you checked\n```\n\n## Troubleshooting\n\n### Memory search not working?\n1. Check `memorySearch.enabled: true` in config\n2. Verify MEMORY.md exists in workspace root\n3. Restart gateway: `clawdbot gateway restart`\n\n### Results not relevant?\n- Lower `minScore` to `0.2` for more results\n- Increase `maxResults` to `30`\n- Check that memory files have meaningful content\n\n### Provider errors?\n- Voyage: Set `VOYAGE_API_KEY` in environment\n- OpenAI: Set `OPENAI_API_KEY` in environment\n- Use `local` provider if no API keys available\n\n## Verification\n\nTest memory is working:\n\n```\nUser: \"What do you remember about [past topic]?\"\nAgent: [Should search memory and return relevant context]\n```\n\nIf agent has no memory, config isn't applied. Restart gateway.\n\n## Full Config Example\n\n```json\n{\n  \"memorySearch\": {\n    \"enabled\": true,\n    \"provider\": \"voyage\",\n    \"sources\": [\"memory\", \"sessions\"],\n    \"indexMode\": \"hot\",\n    \"minScore\": 0.3,\n    \"maxResults\": 20\n  },\n  \"workspace\": \"/path/to/your/workspace\"\n}\n```\n\n## Why This Matters\n\nWithout memory:\n- Agent forgets everything between sessions\n- Repeats questions, loses context\n- No continuity on projects\n\nWith memory:\n- Recalls past conversations\n- Knows your preferences\n- Tracks project history\n- Builds relationship over time\n\nGoldfish → Elephant. 🐘\n","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":743,"installsAllTime":28,"installsCurrent":28,"stars":0,"versions":1},"createdAt":1772742088233,"updatedAt":1778491742347},"latestVersion":{"version":"1.0.0","createdAt":1772742088233,"changelog":"Initial release — enables persistent memory setup and search for Moltbot/Clawdbot agents.\n\n- Provides step-by-step instructions for enabling and configuring memory search in config files.\n- Details recommended workspace memory structure, including MEMORY.md and daily logs.\n- Explains core config options and embedding providers.\n- Offers troubleshooting tips for common memory setup issues.\n- Includes verification steps and best practices for effective agent memory use.","license":null},"metadata":null,"owner":{"handle":"gwsq","userId":"s17dmbwvt7e2dtzv4mpw3n13rx83hk3f","displayName":"Gwsq","image":"https://avatars.githubusercontent.com/u/150670216?v=4"},"moderation":null}