{
  "schemaVersion": "1.0",
  "item": {
    "slug": "openclaw-mem",
    "name": "OpenClaw Memory",
    "source": "tencent",
    "type": "skill",
    "category": "AI 智能",
    "sourceUrl": "https://clawhub.ai/WeAreAllSatoshiN/openclaw-mem",
    "canonicalUrl": "https://clawhub.ai/WeAreAllSatoshiN/openclaw-mem",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "redirect",
    "downloadUrl": "/downloads/openclaw-mem",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=openclaw-mem",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "SKILL.md"
    ],
    "primaryDoc": "SKILL.md",
    "quickSetup": [
      "Download the package from Yavira.",
      "Extract the archive and review SKILL.md first.",
      "Import or place the package into your OpenClaw setup."
    ],
    "agentAssist": {
      "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
      "steps": [
        "Download the package from Yavira.",
        "Extract it into a folder your agent can access.",
        "Paste one of the prompts below and point your agent at the extracted folder."
      ],
      "prompts": [
        {
          "label": "New install",
          "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
        },
        {
          "label": "Upgrade existing",
          "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
        }
      ]
    },
    "sourceHealth": {
      "source": "tencent",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-04-30T16:55:25.780Z",
      "expiresAt": "2026-05-07T16:55:25.780Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network",
        "contentDisposition": "attachment; filename=\"network-1.0.0.zip\"",
        "redirectLocation": null,
        "bodySnippet": null
      },
      "scope": "source",
      "summary": "Source download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this source.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/openclaw-mem"
    },
    "validation": {
      "installChecklist": [
        "Use the Yavira download entry.",
        "Review SKILL.md after the package is downloaded.",
        "Confirm the extracted package contains the expected setup assets."
      ],
      "postInstallChecks": [
        "Confirm the extracted package includes the expected docs or setup files.",
        "Validate the skill or prompts are available in your target agent workspace.",
        "Capture any manual follow-up steps the agent could not complete."
      ]
    },
    "downloadPageUrl": "https://openagent3.xyz/downloads/openclaw-mem",
    "agentPageUrl": "https://openagent3.xyz/skills/openclaw-mem/agent",
    "manifestUrl": "https://openagent3.xyz/skills/openclaw-mem/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/openclaw-mem/agent.md"
  },
  "agentAssist": {
    "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
    "steps": [
      "Download the package from Yavira.",
      "Extract it into a folder your agent can access.",
      "Paste one of the prompts below and point your agent at the extracted folder."
    ],
    "prompts": [
      {
        "label": "New install",
        "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
      },
      {
        "label": "Upgrade existing",
        "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
      }
    ]
  },
  "documentation": {
    "source": "clawhub",
    "primaryDoc": "SKILL.md",
    "sections": [
      {
        "title": "OpenClaw Memory Curator",
        "body": "A session-first memory system for OpenClaw.\n\nIt exists for one reason:\nimportant knowledge must survive session compaction without bloating the context window."
      },
      {
        "title": "TL;DR (for humans)",
        "body": "Session memory = temporary (RAM)\nDisk = source of truth\nDecisions & preferences → MEMORY.md\nDaily work → memory/YYYY-MM-DD.md\nThis skill saves durable knowledge before compaction\nRetrieval always happens via memory_search → memory_get\n\nIf something matters later, write it to disk.\n\n⚠️ CRITICAL REQUIREMENT\nSession memory indexing must be enabled."
      },
      {
        "title": "Enable Session Memory",
        "body": "CLI\n\nclawdbot config set agents.defaults.memorySearch.experimental.sessionMemory true\n\nJSON\n\n{\n  \"agents\": {\n    \"defaults\": {\n      \"memorySearch\": {\n        \"experimental\": { \"sessionMemory\": true },\n        \"sources\": [\"memory\", \"sessions\"]\n      }\n    }\n  }\n}"
      },
      {
        "title": "Mental Model (read this once)",
        "body": "OpenClaw memory has three layers. Confusion usually comes from mixing them up."
      },
      {
        "title": "1. Session Memory (RAM)",
        "body": "Lives in the current conversation\nAutomatically compacted\nIndexed for retrieval\nNever reliable long-term\n\n👉 Treat as short-term thinking space."
      },
      {
        "title": "2. Daily Logs (memory/YYYY-MM-DD.md)",
        "body": "Append-only\nWhat happened today\nCommands, edits, short-lived issues\n\n👉 Treat as a work log, not a knowledge base."
      },
      {
        "title": "3. Long-Term Memory (MEMORY.md)",
        "body": "Curated\nSmall\nHigh-signal only\nIndexed and retrievable\n\n👉 Treat as facts the agent must not forget."
      },
      {
        "title": "Write to MEMORY.md if it would still be true next week.",
        "body": "Examples:\n\nDecisions\nPreferences\nInvariants\nPolicies"
      },
      {
        "title": "Write to daily logs if it helps understand today.",
        "body": "Examples:\n\nRefactors\nExperiments\nTemporary blockers\n\nIf unsure: write to daily log first, promote later."
      },
      {
        "title": "Pre-Compaction Flush (why this exists)",
        "body": "Before OpenClaw compacts the session, it triggers a silent reminder.\n\nThis skill uses that moment as a Save Game checkpoint."
      },
      {
        "title": "What happens:",
        "body": "Durable knowledge is extracted\nDaily notes are written to today’s log\nDurable items are promoted to MEMORY.md\nAgent replies NO_REPLY (user never sees this)\n\nThis prevents knowledge loss without interrupting you."
      },
      {
        "title": "Durable Memory Format (MEMORY.md)",
        "body": "Use IDs and tags so search works reliably.\n\n## DEC-2026-02-04-01\ntype: decision\narea: memory\n\nDecision:\nSession memory is retrieval-only. Disk is the source of truth.\n\nReason:\nSession compaction is lossy. Disk memory is stable."
      },
      {
        "title": "ID prefixes",
        "body": "DEC – Decisions\nPREF – Preferences\nFACT – Durable facts\nPOLICY – Rules / invariants"
      },
      {
        "title": "Retrieval Strategy (how agents should recall)",
        "body": "Use memory_search (max ~6 results)\nPick the best 1–2 hits\nUse memory_get with line ranges\nInject the minimum text required\n\nThis keeps context small and precise."
      },
      {
        "title": "Agent Playbook (rules for agents)",
        "body": "Prefer disk over RAM\nPrefer MEMORY.md over daily logs for facts\nUse search before asking the user again\nNever copy raw chat into memory\nWrite memory explicitly, do not assume it sticks"
      },
      {
        "title": "Anti-Patterns (do not do these)",
        "body": "❌ Copy chat transcripts into memory\n❌ Store secrets or credentials\n❌ Treat daily logs as long-term memory\n❌ Overwrite memory files instead of appending\n❌ Store speculation as fact"
      },
      {
        "title": "Privacy Rules",
        "body": "Never store secrets (API keys, tokens, passwords)\nIgnore anything inside <private>...</private>\nIf sensitive info exists: store only that it exists, not the value"
      },
      {
        "title": "Retention & Cleanup",
        "body": "Default: no deletion\n\nDisk is cheap\nRecall quality is expensive\n\nOptional:\n\nMove old daily logs to memory/archive/YYYY-MM/\nOnly prune after durable knowledge is verified"
      },
      {
        "title": "Usage (human-friendly)",
        "body": "Examples that work well:\n\n“Store this as a durable decision.”\n“This is a preference, remember it.”\n“Write this to today’s log.”"
      },
      {
        "title": "Design Philosophy",
        "body": "Disk is truth\nRAM is convenience\nRetrieval beats retention\nFewer tokens > more tokens\nMemory should earn its place"
      }
    ],
    "body": "OpenClaw Memory Curator\n\nA session-first memory system for OpenClaw.\n\nIt exists for one reason: important knowledge must survive session compaction without bloating the context window.\n\nTL;DR (for humans)\nSession memory = temporary (RAM)\nDisk = source of truth\nDecisions & preferences → MEMORY.md\nDaily work → memory/YYYY-MM-DD.md\nThis skill saves durable knowledge before compaction\nRetrieval always happens via memory_search → memory_get\n\nIf something matters later, write it to disk.\n\n⚠️ CRITICAL REQUIREMENT\n\nSession memory indexing must be enabled.\n\nEnable Session Memory\n\nCLI\n\nclawdbot config set agents.defaults.memorySearch.experimental.sessionMemory true\n\n\nJSON\n\n{\n  \"agents\": {\n    \"defaults\": {\n      \"memorySearch\": {\n        \"experimental\": { \"sessionMemory\": true },\n        \"sources\": [\"memory\", \"sessions\"]\n      }\n    }\n  }\n}\n\nMental Model (read this once)\n\nOpenClaw memory has three layers. Confusion usually comes from mixing them up.\n\n1. Session Memory (RAM)\nLives in the current conversation\nAutomatically compacted\nIndexed for retrieval\nNever reliable long-term\n\n👉 Treat as short-term thinking space.\n\n2. Daily Logs (memory/YYYY-MM-DD.md)\nAppend-only\nWhat happened today\nCommands, edits, short-lived issues\n\n👉 Treat as a work log, not a knowledge base.\n\n3. Long-Term Memory (MEMORY.md)\nCurated\nSmall\nHigh-signal only\nIndexed and retrievable\n\n👉 Treat as facts the agent must not forget.\n\nWhen to Write Memory (simple rules)\nWrite to MEMORY.md if it would still be true next week.\n\nExamples:\n\nDecisions\nPreferences\nInvariants\nPolicies\nWrite to daily logs if it helps understand today.\n\nExamples:\n\nRefactors\nExperiments\nTemporary blockers\n\nIf unsure: write to daily log first, promote later.\n\nPre-Compaction Flush (why this exists)\n\nBefore OpenClaw compacts the session, it triggers a silent reminder.\n\nThis skill uses that moment as a Save Game checkpoint.\n\nWhat happens:\nDurable knowledge is extracted\nDaily notes are written to today’s log\nDurable items are promoted to MEMORY.md\nAgent replies NO_REPLY (user never sees this)\n\nThis prevents knowledge loss without interrupting you.\n\nDurable Memory Format (MEMORY.md)\n\nUse IDs and tags so search works reliably.\n\n## DEC-2026-02-04-01\ntype: decision\narea: memory\n\nDecision:\nSession memory is retrieval-only. Disk is the source of truth.\n\nReason:\nSession compaction is lossy. Disk memory is stable.\n\nID prefixes\nDEC – Decisions\nPREF – Preferences\nFACT – Durable facts\nPOLICY – Rules / invariants\nRetrieval Strategy (how agents should recall)\nUse memory_search (max ~6 results)\nPick the best 1–2 hits\nUse memory_get with line ranges\nInject the minimum text required\n\nThis keeps context small and precise.\n\nAgent Playbook (rules for agents)\nPrefer disk over RAM\nPrefer MEMORY.md over daily logs for facts\nUse search before asking the user again\nNever copy raw chat into memory\nWrite memory explicitly, do not assume it sticks\nAnti-Patterns (do not do these)\n❌ Copy chat transcripts into memory\n❌ Store secrets or credentials\n❌ Treat daily logs as long-term memory\n❌ Overwrite memory files instead of appending\n❌ Store speculation as fact\nPrivacy Rules\nNever store secrets (API keys, tokens, passwords)\nIgnore anything inside <private>...</private>\nIf sensitive info exists: store only that it exists, not the value\nRetention & Cleanup\n\nDefault: no deletion\n\nDisk is cheap\nRecall quality is expensive\n\nOptional:\n\nMove old daily logs to memory/archive/YYYY-MM/\nOnly prune after durable knowledge is verified\nUsage (human-friendly)\n\nExamples that work well:\n\n“Store this as a durable decision.”\n“This is a preference, remember it.”\n“Write this to today’s log.”\nDesign Philosophy\nDisk is truth\nRAM is convenience\nRetrieval beats retention\nFewer tokens > more tokens\nMemory should earn its place"
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/WeAreAllSatoshiN/openclaw-mem",
    "publisherUrl": "https://clawhub.ai/WeAreAllSatoshiN/openclaw-mem",
    "owner": "WeAreAllSatoshiN",
    "version": "2.1.0",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/openclaw-mem",
    "downloadUrl": "https://openagent3.xyz/downloads/openclaw-mem",
    "agentUrl": "https://openagent3.xyz/skills/openclaw-mem/agent",
    "manifestUrl": "https://openagent3.xyz/skills/openclaw-mem/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/openclaw-mem/agent.md"
  }
}