← 返回 Skills 市场
108
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install synap
功能描述
Sovereign AI knowledge infrastructure. Typed entity graph, documents, long-term memory, messaging relay, and AI governance — all in PostgreSQL.
安全使用建议
This skill is coherent: it connects your agent to a Synap pod and uses the SYNAP_HUB_API_KEY to read and write persistent memory and entities on that pod. Before installing, confirm you trust the pod at SYNAP_POD_URL (or self-host using the README steps). Limit the API key scopes (only hub-protocol.read and hub-protocol.write as required), store the key securely, and avoid using a managed/third-party pod for highly sensitive data unless you have strong trust and auditing. Note: the skill can be invoked autonomously by agents (normal for memory skills), so an attacker who gains the key+pod could read/write persistent agent data — rotate keys and monitor audit logs if possible.
功能分析
Type: OpenClaw Skill
Name: synap
Version: 1.0.0
The synap skill provides a structured interface for an AI agent to interact with a knowledge management service via REST API. The SKILL.md file includes explicit safety instructions that prohibit the agent from accessing sensitive local files (e.g., .env, system paths) and defines a governance model requiring user approval for entity and document creation. The behavior is well-documented and aligned with the stated purpose of providing sovereign AI knowledge infrastructure.
能力评估
Purpose & Capability
Name/description match the declared env vars and API endpoints. Requesting SYNAP_HUB_API_KEY and SYNAP_POD_URL is appropriate for a remote hub protocol that provides memory, entities, and relays.
Instruction Scope
SKILL.md explicitly confines actions to the Synap hub REST API under {SYNAP_POD_URL}/api/hub and lists the read/write primitives and governance responses. It does not instruct reading unrelated local files, hidden paths, or other unrelated credentials.
Install Mechanism
Instruction-only skill with no install spec or code files, so nothing is written to disk or downloaded by the installer; low install surface.
Credentials
Only two required env vars (hub API key and pod URL) plus optional, documented vars are requested. The primary credential (SYNAP_HUB_API_KEY) reasonably maps to the described read/write operations; the README notes required scopes.
Persistence & Privilege
The skill does not request always: true and is not asking to modify other skills or system-wide settings. It does allow autonomous invocation (default), which is expected for a memory/knowledge skill; that increases blast radius but is consistent with purpose.
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 synap - 安装完成后，直接呼叫该 Skill 的名称或使用
/synap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
Synap 1.0.0 — Initial Release
- Introduces sovereign AI knowledge infrastructure with a typed entity graph, long-term memory, messaging relay, and AI governance — all backed by PostgreSQL.
- Provides structured memory with search and recall across entities, documents, and facts using Typesense full-text and pgvector semantic search.
- Supports governance workflows: all writes are reviewed or auto-approved as appropriate to ensure data integrity and user oversight.
- Integrates messaging relay for external channels (Telegram, WhatsApp, Slack), allowing messages to be brought into Synap for AI processing.
- Fully documented data model, API endpoints, setup methods, and response formats included for easy onboarding.
元数据
常见问题
Synap 是什么？
Sovereign AI knowledge infrastructure. Typed entity graph, documents, long-term memory, messaging relay, and AI governance — all in PostgreSQL. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 108 次。
如何安装 Synap？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install synap」即可一键安装，无需额外配置。
Synap 是免费的吗？
是的，Synap 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Synap 支持哪些平台？
Synap 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（macos, linux, windows）。
谁开发了 Synap？
由 Antoine（@antoinesrvt）开发并维护，当前版本 v1.0.0。
推荐 Skills