← 返回 Skills 市场
3041
总下载
3
收藏
8
当前安装
1
版本数
在 OpenClaw 中安装
/install vestige
功能描述
Cognitive memory system using FSRS-6 spaced repetition. Memories fade naturally like human memory. Use for persistent recall across sessions.
使用说明 (SKILL.md)
Vestige Memory Skill
Cognitive memory system based on 130 years of memory research. FSRS-6 spaced repetition, spreading activation, synaptic tagging—all running 100% local.
Binary Location
~/bin/vestige-mcp
~/bin/vestige
~/bin/vestige-restore
When to Use
- Persistent memory across sessions
- User preferences ("I prefer TypeScript", "I always use dark mode")
- Bug fixes and solutions worth remembering
- Project patterns and architectural decisions
- Reminders and future triggers
Quick Commands
Search Memory
echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"search","arguments":{"query":"user preferences"}}}' | ~/bin/vestige-mcp 2>/dev/null | jq -r '.result.content[0].text // .error.message'
Save Memory (Smart Ingest)
echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"smart_ingest","arguments":{"content":"User prefers Swiss Modern design style for presentations","tags":["preference","design"]}}}' | ~/bin/vestige-mcp 2>/dev/null | jq -r '.result.content[0].text // .error.message'
Simple Ingest
echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"ingest","arguments":{"content":"TKPay Offline project: POC 2 months, MVP 2 months, budget 250K DH","tags":["project","tkpay"]}}}' | ~/bin/vestige-mcp 2>/dev/null | jq -r '.result.content[0].text // .error.message'
Check Stats
~/bin/vestige stats
Health Check
~/bin/vestige health
MCP Tools Available
| Tool | Description |
|---|---|
search |
Unified search (keyword + semantic + hybrid) |
smart_ingest |
Intelligent ingestion with duplicate detection |
ingest |
Simple memory storage |
memory |
Get, delete, or check memory state |
codebase |
Remember patterns and architectural decisions |
intention |
Set reminders and future triggers |
promote_memory |
Mark memory as helpful (strengthens) |
demote_memory |
Mark memory as wrong (weakens) |
Trigger Words
| User Says | Action |
|---|---|
| "Remember this" | smart_ingest immediately |
| "Don't forget" | smart_ingest with high priority |
| "I always..." / "I never..." | Save as preference |
| "I prefer..." / "I like..." | Save as preference |
| "This is important" | smart_ingest + promote_memory |
| "Remind me..." | Create intention |
Session Start Routine
At the start of conversations, search for relevant context:
# Search user preferences
echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"search","arguments":{"query":"user preferences instructions"}}}' | ~/bin/vestige-mcp 2>/dev/null | jq -r '.result.content[0].text'
# Search project context
echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"search","arguments":{"query":"current project context"}}}' | ~/bin/vestige-mcp 2>/dev/null | jq -r '.result.content[0].text'
Helper Script
For easier usage, create ~/bin/vmem:
#!/bin/bash
# Vestige Memory Helper
ACTION=$1
shift
case $ACTION in
search)
echo "{\"jsonrpc\":\"2.0\",\"id\":1,\"method\":\"tools/call\",\"params\":{\"name\":\"search\",\"arguments\":{\"query\":\"$*\"}}}" | ~/bin/vestige-mcp 2>/dev/null | jq -r '.result.content[0].text // .error.message'
;;
save)
echo "{\"jsonrpc\":\"2.0\",\"id\":1,\"method\":\"tools/call\",\"params\":{\"name\":\"smart_ingest\",\"arguments\":{\"content\":\"$*\"}}}" | ~/bin/vestige-mcp 2>/dev/null | jq -r '.result.content[0].text // .error.message'
;;
stats)
~/bin/vestige stats
;;
*)
echo "Usage: vmem [search|save|stats] [content]"
;;
esac
Data Location
- macOS:
~/Library/Application Support/com.vestige.core/ - Linux:
~/.local/share/vestige/ - Embedding cache:
~/Library/Caches/com.vestige.core/fastembed/
Integration Notes
Vestige complements the existing memory/ folder system:
- memory/*.md = Human-readable daily logs
- MEMORY.md = Curated long-term notes
- Vestige = Semantic search + automatic decay + spaced repetition
Use Vestige for:
- Things you want to recall semantically (not just keyword search)
- Preferences that should persist indefinitely
- Solutions worth remembering (with automatic decay if unused)
安全使用建议
This skill is an instruction-only adapter for a local memory service — it will only work if you already have the referenced binaries (~/bin/vestige-mcp, ~/bin/vestige, etc.) and utilities like jq installed. Before enabling: 1) verify you own/trust the vestige binaries and install them from a reputable source (the SKILL.md gives no install or checksum guidance); 2) check and secure the data directories (~/Library/Application Support/com.vestige.core/, ~/.local/share/vestige/) for sensitive or PII content and consider encryption/backup policies; 3) confirm whether you want the agent to run the session-start search commands automatically (they will call local binaries); and 4) ensure jq and bash are available on PATH or adjust the helper script accordingly. The main practical risk is running unknown local binaries — treat them like any third‑party program and inspect/verify them before use.
功能分析
Type: OpenClaw Skill
Name: vestige
Version: 1.0.0
The skill bundle is classified as benign. The `SKILL.md` file primarily defines a cognitive memory system, providing instructions for the AI agent to interact with local binaries (`~/bin/vestige-mcp`, `~/bin/vestige`) for tasks like searching, ingesting, and managing memories. All commands and instructions, including the helper script `~/bin/vmem`, are directly aligned with the stated purpose of a memory system. There is no evidence of data exfiltration, malicious execution of remote payloads, persistence mechanisms, or prompt injection attempts designed to subvert the agent's core function or access sensitive data beyond the skill's scope.
能力评估
Purpose & Capability
The SKILL.md describes a local FSRS-based memory system and the runtime examples call local binaries (~/bin/vestige-mcp, ~/bin/vestige, ~/bin/vestige-restore) and reference user-data paths; this is coherent with the stated purpose. However, the registry metadata declared no required binaries even though the instructions clearly require specific executables and common tools (jq, bash) to be present.
Instruction Scope
Instructions focus on searching, ingesting, and promoting memories via local CLI tools and helper scripts; they do not request unrelated system data, network endpoints, or extra credentials. They do instruct the agent to run local binaries at session start (searches), which is consistent with providing persistent memory context.
Install Mechanism
There is no install spec (instruction-only), so nothing is written or installed by the skill itself. That lowers installer risk but means the skill assumes external binaries are already installed; the SKILL.md provides no safe download or verification guidance for those binaries.
Credentials
No environment variables, credentials, or external service tokens are requested. Data locations are local to the user's home directory and consistent with a local-memory design.
Persistence & Privilege
The skill does not request always: true and defaults to normal invocation behavior. It does propose a session-start routine that will call local tools for context, which is reasonable for a memory plugin but something to be aware of if you prefer to limit autonomous calls.
如何使用
- 确保已安装 OpenClaw(本地或 Docker 部署)
- 在对话框中输入安装命令:
/install vestige - 安装完成后,直接呼叫该 Skill 的名称或使用
/vestige触发 - 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
Initial release - Cognitive memory system using FSRS-6 spaced repetition
元数据
常见问题
Vestige 是什么?
Cognitive memory system using FSRS-6 spaced repetition. Memories fade naturally like human memory. Use for persistent recall across sessions. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 3041 次。
如何安装 Vestige?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install vestige」即可一键安装,无需额外配置。
Vestige 是免费的吗?
是的,Vestige 完全免费(开源免费),可自由下载、安装和使用。
Vestige 支持哪些平台?
Vestige 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。
谁开发了 Vestige?
由 Belkouche(@belkouche)开发并维护,当前版本 v1.0.0。
推荐 Skills