Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.
Python agent-contextagent-memoryagent-pluginsagentic-ragcontext-databasedsh-pluginknowledge-basememorymemory-managementself-evolvingvector-database
docs: add general agent integration guides (#2354)
yufeng5752ceb8fb20
diff --git a/.github/workflows/docs-tos.yml b/.github/workflows/docs-tos.ymlindex f45139a29..2d66e098c 100644--- a/.github/workflows/docs-tos.yml+++ b/.github/workflows/docs-tos.yml@@ -134,14 +134,15 @@ jobs:--cache-control "${NO_CACHE_CONTROL}" \--only-show-errors- if [ -f "${DIST_DIR}/agents/index.json" ]; then- aws s3 cp "${DIST_DIR}/agents/index.json" "s3://${TOS_BUCKET}/agents/index.json" \+ while IFS= read -r -d '' json_file; do+ rel_path="${json_file#${DIST_DIR}/agents/}"+ aws s3 cp "${json_file}" "s3://${TOS_BUCKET}/agents/${rel_path}" \--endpoint-url "https://${TOS_S3_ENDPOINT}" \--content-type "application/json; charset=utf-8" \--content-disposition "inline" \--cache-control "${NO_CACHE_CONTROL}" \--only-show-errors- fi+ done < <(find "${DIST_DIR}/agents" -type f -name "*.json" -print0)while IFS= read -r -d '' markdown_file; dorel_path="${markdown_file#${DIST_DIR}/agents/}"diff --git a/docs/images/agents/en/api.md b/docs/images/agents/en/api.mdnew file mode 100644index 000000000..8ac426e83--- /dev/null+++ b/docs/images/agents/en/api.md@@ -0,0 +1,86 @@+# API Integration++## Step 1: Choose an API Key++Copy the API Key shown in the OpenViking console and pass it to your agent environment.++## Step 2: Add a resource++Use the REST API to upload a local file and create a resource:++```python+import json+from pathlib import Path++import requests++url = "https://api.vikingdb.cn-beijing.volces.com/openviking"+api_key = "<your-api-key>"+agent_id = "<your-agent-id>"+file_path = Path("<your-file-path>")+resource_to = "viking://resources/test.txt"+reason = "External API documentation"++headers = {+ "Content-Type": "application/json",+ "Authorization": "Bearer " + api_key,+ "X-OpenViking-Agent": agent_id,+}++def post_json(path: str, payload: dict, timeout: float):+ response = requests.post(f"{url}{path}", headers=headers, json=payload, timeout=timeout)+ response.raise_for_status()+ return response.json()++with file_path.open("rb") as file:+ result = requests.post(+ f"{url}/api/v1/resources/temp_upload",+ headers={+ "Authorization": "Bearer " + api_key,+ "X-OpenViking-Agent": agent_id,+ },+ files={"file": (file_path.name, file, "application/octet-stream")},+ timeout=120.0,+ )+result.raise_for_status()+temp_file_id = result.json()["result"]["temp_file_id"]++result = post_json(+ "/api/v1/resources",+ {+ "temp_file_id": temp_file_id,+ "source_name": file_path.name,+ "to": resource_to,+ "reason": reason,+ },+ 120.0,+)+print(json.dumps(result, ensure_ascii=False, indent=2))+```++## Step 3: Add memory++Create a session, add a message, and commit the session:++```python+text = "I am a developer"++session = post_json("/api/v1/sessions", {}, 360.0)+session_id = session["result"]["session_id"]++post_json(+ f"/api/v1/sessions/{session_id}/messages",+ {+ "role": "user",+ "parts": [{"type": "text", "text": text}],+ },+ 360.0,+)++result = post_json(+ f"/api/v1/sessions/{session_id}/commit",+ {"telemetry": False},+ 360.0,+)+print(json.dumps(result, ensure_ascii=False, indent=2))+```diff --git a/docs/images/agents/en/claude-code.md b/docs/images/agents/en/claude-code.mdnew file mode 100644index 000000000..b662ab1ad--- /dev/null+++ b/docs/images/agents/en/claude-code.md@@ -0,0 +1,27 @@+# Claude Code Memory Plugin++Add cross-project, cross-session long-term memory to [Claude Code](https://docs.claude.com/en/docs/claude-code/overview). Install once; recall and capture happen automatically on every conversation.++## Install++```bash+bash <(curl -fsSL https://raw.githubusercontent.com/volcengine/OpenViking/main/examples/claude-code-memory-plugin/setup-helper/install.sh)+```++The installer checks dependencies, configures the OpenViking connection, and installs the plugin.++## Verify++```bash+type claude+```++Expected output: `claude is a shell function`.++In Claude Code:++- `/plugins`: find `openviking-memory` in Installed+- `/mcp`: confirm the OpenViking MCP endpoint is connected+- `/openviking-memory:ov`: view server status, identity, recall, and capture state++Set `OPENVIKING_DEBUG=1` to write debug logs to `~/.openviking/logs/cc-hooks.log`.diff --git a/docs/images/agents/en/cli.md b/docs/images/agents/en/cli.mdnew file mode 100644index 000000000..ca902820b--- /dev/null+++ b/docs/images/agents/en/cli.md@@ -0,0 +1,40 @@+# CLI Integration++## Install from the command line++Run the following command in your terminal to install OpenViking CLI and start configuration:++```bash+npm i -g @openviking/cli && ov config+```++Use this service URL when prompted:++```text+https://api.vikingdb.cn-beijing.volces.com/openviking+```++After configuration, inspect CLI usage with:++```bash+ov --help+```++## Install through an agent conversation++The Agent CLI setup flow will be improved in a follow-up PR. For now, send the following instruction to your agent:++```text+Write the following content to ~/.openviking/ovcli.conf:+{"url":"https://api.vikingdb.cn-beijing.volces.com/openviking","api_key":"Please ask user for Volcengine OpenViking API key."}++Ask the user for the API Key. If ~/.openviking/ovcli.conf already exists and the content conflicts, ask the user whether to back up the original file before overwriting it.++Install OpenViking CLI:+npm i -g @openviking/cli++After installation, run:+ov --help++Explore the CLI usage and write the OpenViking CLI workflow into your long-term memory.+```diff --git a/docs/images/agents/en/codex.md b/docs/images/agents/en/codex.mdnew file mode 100644index 000000000..b19f314e7--- /dev/null+++ b/docs/images/agents/en/codex.md@@ -0,0 +1,32 @@+# Codex Memory Plugin++Add persistent cross-session memory to Codex. The plugin recalls relevant memory before each user prompt, captures conversation updates after each turn, commits before compaction, and connects Codex to the OpenViking MCP endpoint.++## Install++```bash+bash <(curl -fsSL https://raw.githubusercontent.com/volcengine/OpenViking/main/examples/codex-memory-plugin/setup-helper/install.sh)+```++After installation:++```bash+source ~/.zshrc+codex+```++Approve hooks on first launch if Codex asks for review.++## Verify++```bash+type codex+```++Expected output: `codex is a shell function`.++Set `OPENVIKING_DEBUG=1` to write events to `~/.openviking/logs/codex-hooks.log`.++## Notes++Codex does not trigger hooks on `SIGTERM`, `Ctrl+C`, or `/exit`. The plugin cleans up orphan sessions on the next `SessionStart` with an idle TTL.diff --git a/docs/images/agents/en/cursor.md b/docs/images/agents/en/cursor.mdnew file mode 100644index 000000000..927e5b6e0--- /dev/null+++ b/docs/images/agents/en/cursor.md@@ -0,0 +1,49 @@+# Cursor MCP Integration++## Step 1: Get an API Key++OpenViking MCP clients use the API Key from the OpenViking console as the authorization token. Copy it before configuring Cursor.++## Step 2: Open Cursor MCP settings++In Cursor, open Settings and go to **Tools & MCPs**.++## Step 3: Add a custom MCP server++Click **Add Custom MCP**, then add the following configuration to `mcp.json`:++```json+{+ "mcpServers": {+ "ov-mcp-server": {+ "url": "https://api.vikingdb.cn-beijing.volces.com/openviking/mcp",+ "headers": {+ "Authorization": "Bearer <API Key>"+ }+ }+ }+}+```++The `Authorization` value must include the `Bearer` prefix.++## Step 4: Enable and verify++Save `mcp.json`. Cursor should load `ov-mcp-server` and show the enabled tools.++Run these queries in Cursor:++```text+ov ls+ov health+```++`ov ls` should return OpenViking root directories such as `agent`, `resources`, `session`, and `user`. `ov health` should return service status and the current identity.++## Troubleshooting++| Problem | Fix |+|---|---|+| `401 Unauthorized` | Check that `Authorization` includes `Bearer <API Key>` and that the API Key is valid. |+| Network timeout | Confirm the network can reach `api.vikingdb.cn-beijing.volces.com`. |+| Agent cannot see tools | Confirm the MCP server is enabled. Some clients need a restart after config changes. |diff --git a/docs/images/agents/en/hermes.md b/docs/images/agents/en/hermes.mdnew file mode 100644index 000000000..149f5e425--- /dev/null+++ b/docs/images/agents/en/hermes.md@@ -0,0 +1,27 @@+# Hermes Agent++[Hermes Agent](https://hermes-agent.nousresearch.com/) includes OpenViking as a native memory provider. No plugin installation is required.++## Configure memory++Run the Hermes memory setup wizard:++```bash+hermes memory setup+```++The wizard asks for:++- OpenViking service URL, such as `https://api.vikingdb.cn-beijing.volces.com/openviking`+- API Key+- Tenant account, user, and agent ID for multi-tenant deployments++Hermes stores the configuration in `config.yaml` and `.env`.++## Verify++```bash+hermes memory status+```++After setup, Hermes uses OpenViking for long-term memory storage, recall, and extraction.diff --git a/docs/images/agents/en/index.json b/docs/images/agents/en/index.jsonnew file mode 100644index 000000000..2799db3a8--- /dev/null+++ b/docs/images/agents/en/index.json@@ -0,0 +1,93 @@+{+ "version": 1,+ "updatedAt": "2026-06-01",+ "cdnBaseUrl": "https://docs.openviking.net/",+ "accessMethods": [+ {+ "id": "mcp",+ "name": "MCP Integration",+ "icon": "link",+ "summary": "Connect from any client that supports MCP.",+ "detailPath": "https://docs.openviking.net/agents/en/mcp.md"+ },+ {+ "id": "cli",+ "name": "CLI Integration",+ "icon": "terminal",+ "summary": "Connect agents through the OpenViking command line.",+ "detailPath": "https://docs.openviking.net/agents/en/cli.md"+ },+ {+ "id": "api",+ "name": "API Integration",+ "icon": "code",+ "summary": "Call OpenViking capabilities through the REST API.",+ "detailPath": "https://docs.openviking.net/agents/en/api.md"+ },+ {+ "id": "sdk",+ "name": "SDK Integration",+ "icon": "paperclip",+ "summary": "Use SDK wrappers for faster and cleaner integration.",+ "detailPath": "https://docs.openviking.net/agents/en/sdk.md"+ }+ ],+ "agents": [+ {+ "id": "openclaw",+ "name": "OpenClaw",+ "tags": ["Plugin"],+ "logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/OpenClaw.png",+ "summary": "Plugin integration for OpenClaw as a context engine, with cross-agent memory and one-command setup through the npm helper.",+ "detailPath": "https://docs.openviking.net/agents/en/openclaw.md"+ },+ {+ "id": "hermes",+ "name": "Hermes",+ "tags": ["Provider"],+ "logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/Hermes-Agent.png",+ "summary": "Hermes includes OpenViking as a native memory provider. Point Hermes to your OpenViking service to enable storage, recall, and extraction.",+ "detailPath": "https://docs.openviking.net/agents/en/hermes.md"+ },+ {+ "id": "claude-code",+ "name": "Claude Code",+ "tags": ["Plugin"],+ "logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/Claude-Code.png",+ "summary": "Adds long-term memory to Claude Code across projects and sessions. Install once; recall and capture run automatically.",+ "detailPath": "https://docs.openviking.net/agents/en/claude-code.md"+ },+ {+ "id": "codex",+ "name": "Codex",+ "tags": ["Plugin"],+ "logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/codex.png",+ "summary": "Adds persistent cross-session memory to Codex through lifecycle hooks and the OpenViking MCP endpoint.",+ "detailPath": "https://docs.openviking.net/agents/en/codex.md"+ },+ {+ "id": "cursor",+ "name": "Cursor",+ "tags": ["MCP"],+ "logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/Cursor.png",+ "summary": "Connect Cursor through MCP to provide coding and agent-development context management.",+ "detailPath": "https://docs.openviking.net/agents/en/cursor.md"+ },+ {+ "id": "trae",+ "name": "TRAE",+ "tags": ["MCP"],+ "logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/Trae.png",+ "summary": "Connect TRAE through MCP and use OpenViking for coding context, memory, and resource management.",+ "detailPath": "https://docs.openviking.net/agents/en/trae.md"+ },+ {+ "id": "opencode",+ "name": "Opencode",+ "tags": ["Plugin"],+ "logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/OpenCode.png",+ "summary": "Provides OpenCode plugins that can expose OpenViking either as model-selected tools or injected context.",+ "detailPath": "https://docs.openviking.net/agents/en/opencode.md"+ }+ ]+}diff --git a/docs/images/agents/en/mcp.md b/docs/images/agents/en/mcp.mdnew file mode 100644index 000000000..2f81528cd--- /dev/null+++ b/docs/images/agents/en/mcp.md@@ -0,0 +1,30 @@+# MCP Integration++## Step 1: Configure MCP++Add the OpenViking MCP server to your MCP client configuration:++```json+{+ "mcpServers": {+ "ov-mcp-server": {+ "url": "https://api.vikingdb.cn-beijing.volces.com/openviking/mcp",+ "headers": {+ "Authorization": "Bearer <API Key>"+ }+ }+ }+}+```++The `Authorization` value must include the `Bearer` prefix and one space before the API key.++## Step 2: Verify the connection++Ask your agent to run:++```text+ov health+```++If the tool returns the OpenViking version and service status, the MCP connection is ready.diff --git a/docs/images/agents/en/openclaw.md b/docs/images/agents/en/openclaw.mdnew file mode 100644index 000000000..7a1a044dd--- /dev/null+++ b/docs/images/agents/en/openclaw.md@@ -0,0 +1,24 @@+# OpenClaw Plugin++## Step 1: Install OpenViking++Run the following command in the terminal where OpenClaw is installed:++```bash+openclaw plugins install clawhub:@openviking/openclaw-plugin && openclaw openviking setup+```++## Step 2: Enter connection information++The setup command will prompt for:++- Base URL: `https://api.vikingdb.cn-beijing.volces.com/openviking`+- API Key: copy the API Key shown in the OpenViking console++## Step 3: Restart OpenClaw++Restart OpenClaw so the plugin can load the new configuration:++```bash+openclaw gateway restart+```diff --git a/docs/images/agents/en/opencode.md b/docs/images/agents/en/opencode.mdnew file mode 100644index 000000000..498036341--- /dev/null+++ b/docs/images/agents/en/opencode.md@@ -0,0 +1,18 @@+# Opencode Plugin++OpenViking provides two OpenCode plugin modes:++- tool mode, where the model can choose when to call OpenViking tools+- context mode, where OpenViking injects relevant context into the session++## Configure++Install the OpenViking OpenCode plugin from the OpenViking plugin package or local marketplace, then configure the service URL and API Key:++```text+https://api.vikingdb.cn-beijing.volces.com/openviking+```++## Verify++Start OpenCode and confirm the OpenViking plugin is loaded. Ask the agent to inspect OpenViking status or search memory to verify connectivity.diff --git a/docs/images/agents/en/sdk.md b/docs/images/agents/en/sdk.mdnew file mode 100644index 000000000..150032e57--- /dev/null+++ b/docs/images/agents/en/sdk.md@@ -0,0 +1,60 @@+# SDK Integration++## Step 1: Install OpenViking++Install or upgrade the Python package:++```bash+pip install openviking --upgrade --force-reinstall+```++## Step 2: Initialize the client++```python+from openviking.client import SyncHTTPClient++url = "https://api.vikingdb.cn-beijing.volces.com/openviking"+api_key = "<your-api-key>"+agent_id = "<your-agent-id>"++client = SyncHTTPClient(+ url=url,+ api_key=api_key,+ agent_id=agent_id,+ timeout=120.0,+)+client.initialize()+```++## Step 3: Choose an API Key++Copy the API Key shown in the OpenViking console and pass it to the SDK client.++## Step 4: Add a resource++```python+file_path = "<your-file-path>"+resource_to = "viking://resources"+reason = "External API documentation"++client.add_resource(+ path=file_path,+ to=resource_to,+ reason=reason,+)+```++## Step 5: Add memory++```python+text = "I am a developer"++session = client.create_session()+session_id = session["session_id"]+client.add_message(+ session_id,+ "user",+ parts=[{"type": "text", "text": text}],+)+result = client.commit_session(session_id)+```diff --git a/docs/images/agents/en/trae.md b/docs/images/agents/en/trae.mdnew file mode 100644index 000000000..0a08a44d1--- /dev/null+++ b/docs/images/agents/en/trae.md@@ -0,0 +1,47 @@+# TRAE MCP Integration++## Step 1: Get an API Key++Copy the API Key from the OpenViking console. It will be used as the MCP `Authorization` token.++## Step 2: Open TRAE MCP settings++In TRAE, open Settings and select **MCP** to enter the MCP Servers page.++## Step 3: Add an MCP server++Choose manual configuration and paste:++```json+{+ "mcpServers": {+ "ov-mcp-server": {+ "url": "https://api.vikingdb.cn-beijing.volces.com/openviking/mcp",+ "headers": {+ "Authorization": "Bearer <API Key>"+ }+ }+ }+}+```++## Step 4: Enable and verify++After saving, TRAE should load `ov-mcp-server` and enable the tools.++Run:++```text+ov ls+ov health+```++The integration is ready when `ov ls` returns OpenViking directories and `ov health` returns service status.++## Troubleshooting++| Problem | Fix |+|---|---|+| `401 Unauthorized` | Check the API Key and the `Bearer` prefix. |+| Network timeout | Confirm the network can reach the OpenViking API domain. |+| Tool schema is incompatible with the current model | Switch models or upgrade TRAE to the latest version. |diff --git a/docs/images/agents/zh/api.md b/docs/images/agents/zh/api.mdnew file mode 100644index 000000000..c16d2156f--- /dev/null+++ b/docs/images/agents/zh/api.md@@ -0,0 +1,95 @@+## 步骤 1 选择 API Key+复制页面中展示的 API Key 到你的 Agent 终端++## 步骤 2 写入资源+参考 GitHub 提供的资源写入示例,自动填入 API Key 和域名++```python+import json+from pathlib import Path++import requests++url = " `https://api.vikingdb.cn-beijing.volces.com/openviking` "+api_key = "[TODO]your-api-key"+agent_id = "[TODO]your-agent-id"+file_path = Path("[TODO]your-file-path") # e.g. test.txt+resource_to = "[TODO]your-resource-path" # e.g. viking://resources/test.txt+reason = "[TODO]your-reason" # e.g. External API documentation++# 1. Initialize request headers.+headers = {+ "Content-Type": "application/json",+ "Authorization": "Bearer " + api_key,+ "X-OpenViking-Agent": agent_id,+}++def post_json(path: str, payload: dict, timeout: float):+ response = requests.post(f"{url}{path}", headers=headers, json=payload, timeout=timeout)+ response.raise_for_status()+ return response.json()+++# 2. Upload the local file to a temporary resource.+with file_path.open("rb") as file:+ result = requests.post(+ f"{url}/api/v1/resources/temp_upload",+ headers={+ "Authorization": "Bearer " + api_key,+ "X-OpenViking-Agent": agent_id,+ },+ files={"file": (file_path.name, file, "application/octet-stream")},+ timeout=120.0,+ )+result.raise_for_status()+result = result.json()+print(json.dumps(result, ensure_ascii=False, indent=2))+temp_file_id = result["result"]["temp_file_id"]++# 3. Create a resource from the temporary file.+result = post_json(+ "/api/v1/resources",+ {+ "temp_file_id": temp_file_id,+ "source_name": file_path.name,+ "to": resource_to,+ "reason": reason,+ },+ 120.0,+)+print(json.dumps(result, ensure_ascii=False, indent=2))+```++## 步骤 3 写入记忆+参考 GitHub 提供的记忆写入示例,自动填入 API Key 和域名++```python+text = "[TODO]your-message-text" # e.g. I am a developer++def post_json(path: str, payload: dict, timeout: float):+ response = requests.post(f"{url}{path}", headers=headers, json=payload, timeout=timeout)+ response.raise_for_status()+ return response.json()++# Create a session.+session = post_json("/api/v1/sessions", {}, 360.0)+session_id = session["result"]["session_id"]++# Add a message.+post_json(+ f"/api/v1/sessions/{session_id}/messages",+ {+ "role": "user",+ "parts": [{"type": "text", "text": text}],+ },+ 360.0,+)++# Commit the session.+result = post_json(+ f"/api/v1/sessions/{session_id}/commit",+ {"telemetry": False},+ 360.0,+)+print(json.dumps(result, ensure_ascii=False, indent=2))+```diff --git a/docs/images/agents/claude-code.md b/docs/images/agents/zh/claude-code.mdsimilarity index 100%rename from docs/images/agents/claude-code.mdrename to docs/images/agents/zh/claude-code.mddiff --git a/docs/images/agents/zh/cli.md b/docs/images/agents/zh/cli.mdnew file mode 100644index 000000000..f537d8b86--- /dev/null+++ b/docs/images/agents/zh/cli.md@@ -0,0 +1,40 @@+# CLI 接入++## CLI 命令行安装++在终端执行以下命令安装 OpenViking CLI,并进入配置流程:++```bash+npm i -g @openviking/cli && ov config+```++按提示填写 OpenViking 服务地址和 API Key。服务地址可使用:++```text+https://api.vikingdb.cn-beijing.volces.com/openviking+```++配置完成后,可运行以下命令查看 CLI 用法:++```bash+ov --help+```++## Agent 对话安装++下个 PR 将继续优化 Agent CLI 安装流程,目前可先把以下指令发送给你的 Agent,让 Agent 帮你完成安装和配置:++```text+请在 ~/.openviking/ovcli.conf 写入以下内容:+{"url":"https://api.vikingdb.cn-beijing.volces.com/openviking","api_key":"Please ask user for Volcengine OpenViking API key."}++请向用户询问 API Key。如发现 ~/.openviking/ovcli.conf 已存在且内容冲突,请先询问用户是否备份原文件,并在得到确认后再覆盖。++请安装 OpenViking CLI:+npm i -g @openviking/cli++安装完成后,请运行:+ov --help++请探索 CLI 用法,并把 OpenViking CLI 的使用方式写入你的长期记忆。+```diff --git a/docs/images/agents/codex.md b/docs/images/agents/zh/codex.mdsimilarity index 100%rename from docs/images/agents/codex.mdrename to docs/images/agents/zh/codex.mddiff --git a/docs/images/agents/cursor.md b/docs/images/agents/zh/cursor.mdsimilarity index 100%rename from docs/images/agents/cursor.mdrename to docs/images/agents/zh/cursor.mddiff --git a/docs/images/agents/hermes.md b/docs/images/agents/zh/hermes.mdsimilarity index 100%rename from docs/images/agents/hermes.mdrename to docs/images/agents/zh/hermes.mddiff --git a/docs/images/agents/index.json b/docs/images/agents/zh/index.jsonsimilarity index 65%rename from docs/images/agents/index.jsonrename to docs/images/agents/zh/index.jsonindex a260419aa..9ac5886f0 100644--- a/docs/images/agents/index.json+++ b/docs/images/agents/zh/index.json@@ -1,7 +1,37 @@{"version": 1,- "updatedAt": "2026-05-29",+ "updatedAt": "2026-06-01","cdnBaseUrl": "https://docs.openviking.net/",+ "accessMethods": [+ {+ "id": "mcp",+ "name": "MCP 接入",+ "icon": "link",+ "summary": "与任意支持 MCP 的客户端接入",+ "detailPath": "https://docs.openviking.net/agents/zh/mcp.md"+ },+ {+ "id": "cli",+ "name": "CLI 接入",+ "icon": "terminal",+ "summary": "命令行接入 Agent,使用灵活性更高",+ "detailPath": "https://docs.openviking.net/agents/zh/cli.md"+ },+ {+ "id": "api",+ "name": "API 接入",+ "icon": "code",+ "summary": "标准 REST API,灵活调用所有能力",+ "detailPath": "https://docs.openviking.net/agents/zh/api.md"+ },+ {+ "id": "sdk",+ "name": "SDK 接入",+ "icon": "paperclip",+ "summary": "多语言 SDK 封装,集成更便捷、高效",+ "detailPath": "https://docs.openviking.net/agents/zh/sdk.md"+ }+ ],"agents": [{"id": "openclaw",@@ -9,7 +39,7 @@"tags": ["Plugin"],"logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/OpenClaw.png","summary": "Plugin 接入,作为 OpenClaw 的 context engine,支持多 Agent 跨会话记忆与自迭代,通过 npm 安装助手工具一键完成配置",- "detailPath": "https://docs.openviking.net/agents/openclaw.md"+ "detailPath": "https://docs.openviking.net/agents/zh/openclaw.md"},{"id": "hermes",@@ -17,7 +47,7 @@"tags": ["Provider"],"logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/Hermes-Agent.png","summary": "Hermes 官方内置 OpenViking 作为记忆提供方。无需安装插件——直接把 Hermes 指向你的 OpenViking 服务即可,记忆存储、召回和抽取均原生支持",- "detailPath": "https://docs.openviking.net/agents/hermes.md"+ "detailPath": "https://docs.openviking.net/agents/zh/hermes.md"},{"id": "claude-code",@@ -25,7 +55,7 @@"tags": ["Plugin"],"logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/Claude-Code.png","summary": "为 Claude Code 提供跨项目、跨 session 的长期记忆,越用越聪明。安装一次,每次对话自动召回和捕获,模型不需要主动调用任何工具",- "detailPath": "https://docs.openviking.net/agents/claude-code.md"+ "detailPath": "https://docs.openviking.net/agents/zh/claude-code.md"},{"id": "codex",@@ -33,7 +63,7 @@"tags": ["Plugin"],"logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/codex.png","summary": "为 Codex 提供持久化的跨 session 记忆。安装一次——每次用户输入前自动召回,每轮结束后增量捕获,compaction 前提交给记忆抽取器",- "detailPath": "https://docs.openviking.net/agents/codex.md"+ "detailPath": "https://docs.openviking.net/agents/zh/codex.md"},{"id": "cursor",@@ -41,7 +71,7 @@"tags": ["MCP"],"logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/Cursor.png","summary": "通过 MCP 标准协议接入,为 Codex 提供支撑日常编码 和 Agent 开发场景的上下文管理",- "detailPath": "https://docs.openviking.net/agents/cursor.md"+ "detailPath": "https://docs.openviking.net/agents/zh/cursor.md"},{"id": "trae",@@ -49,7 +79,7 @@"tags": ["MCP"],"logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/Trae.png","summary": "通过 MCP 标准协议接入,为 TRAE 提供支撑日常编码 和 Agent 开发场景的上下文管理,可以修改规则和命令来保证更稳定的 MCP 工具使用",- "detailPath": "https://docs.openviking.net/agents/trae.md"+ "detailPath": "https://docs.openviking.net/agents/zh/trae.md"},{"id": "opencode",@@ -57,7 +87,7 @@"tags": ["Plugin"],"logo": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lm_sth/ljhwZthlaukjlkulzlp/agent_logo/OpenCode.png","summary": "提供两种 OpenCode 插件,可以模型自主选择使用,也可以注入上下文使用",- "detailPath": "https://docs.openviking.net/agents/opencode.md"+ "detailPath": "https://docs.openviking.net/agents/zh/opencode.md"}]}diff --git a/docs/images/agents/zh/mcp.md b/docs/images/agents/zh/mcp.mdnew file mode 100644index 000000000..b0bc02ce1--- /dev/null+++ b/docs/images/agents/zh/mcp.md@@ -0,0 +1,22 @@+# MCP 接入++步骤 1 MCP 配置++```json+{+ "mcpServers": {+ "ov-mcp-server": {+ "url": "https://api.vikingdb.cn-beijing.volces.com/openviking/mcp",+ "headers": {+ "Authorization": "Bearer ZGVmYXV********YzdlZjhiMg"+ }+ }+ }+}+```++**关键说明**:`Authorization` 的值需带上 `Bearer` 前缀(注意空格),完整格式为 `Bearer \<API Key\>`。++步骤 2 测试 MCP 工具连通性++输入 `ov health` 检查 ov 的版本和连接状态diff --git a/docs/images/agents/openclaw.md b/docs/images/agents/zh/openclaw.mdsimilarity index 100%rename from docs/images/agents/openclaw.mdrename to docs/images/agents/zh/openclaw.mddiff --git a/docs/images/agents/opencode.md b/docs/images/agents/zh/opencode.mdsimilarity index 100%rename from docs/images/agents/opencode.mdrename to docs/images/agents/zh/opencode.mddiff --git a/docs/images/agents/zh/sdk.md b/docs/images/agents/zh/sdk.mdnew file mode 100644index 000000000..b85e1e5f4--- /dev/null+++ b/docs/images/agents/zh/sdk.md@@ -0,0 +1,61 @@+## 步骤 1 安装 OpenViking+执行以下命令以安装 OpenViking++```bash+pip install openviking --upgrade --force-reinstall+```++## 步骤 2 初始化客户端+参考 GitHub 提供的规范写入示例,自动填入 API Key 和域名++```python+from openviking.client import SyncHTTPClient++url = " `https://api.vikingdb.cn-beijing.volces.com/openviking` "+api_key = "[TODO]your-api-key"+agent_id = "[TODO]your-agent-id"++client = SyncHTTPClient(+ url=url,+ api_key=api_key,+ agent_id=agent_id,+ timeout=120.0,+)+client.initialize()+```++## 步骤 3 选择 API Key+复制页面中展示的 API Key 到你的 Agent 终端++## 步骤 4 写入资源+参考 GitHub 提供的规范写入示例,自动填入 API Key 和域名++```python+file_path = "[TODO]your-file-path"+resource_to = "[TODO]your-resource-path" # e.g. viking://resources+reason = "[TODO]your-reason" # e.g. External API documentation++# Reuse the initialized client.+client.add_resource(+ path=file_path,+ to=resource_to,+ reason=reason,+)+```++## 步骤 5 写入记忆+参考 GitHub 提供的记忆写入示例,自动填入 API Key 和域名++```python+text = "[TODO]your-message-text" # e.g. I am a developer++# Reuse the initialized client.+session = client.create_session()+session_id = session["session_id"]+client.add_message(+ session_id,+ "user",+ parts=[{"type": "text", "text": text}],+)+result = client.commit_session(session_id)+```diff --git a/docs/images/agents/trae.md b/docs/images/agents/zh/trae.mdsimilarity index 99%rename from docs/images/agents/trae.mdrename to docs/images/agents/zh/trae.mdindex 63be76e22..5dc190881 100644--- a/docs/images/agents/trae.md+++ b/docs/images/agents/zh/trae.md@@ -122,7 +122,3 @@|连接失败 / 网络超时<br>|确认网络可访问 `api\.vikingdb\.cn\-beijing\.volces\.com`;企业网络请配置代理白名单||Agent 无法识别工具|检查 MCP Server 是否已\&\#34;启用\&\#34;;部分客户端需重启进程后加载新配置||mcp工具因 argument schema 与当前模型不兼容,请切换/修复 mcp server 或切换模型 \(4027\)|<br><br>尝试切换模型或升级到最新版 Trae|----> (注:内容由 AI 生成,请谨慎参考)