Security audit
Security checks for vulnerabilities and agentic risk
This appears to be a legitimate AIOps integration, but it needs review because it can handle admin and host passwords over unsafe defaults and make persistent platform changes.
Install only for a trusted internal Zhihu AIOps deployment. Require HTTPS for non-local endpoints, use least-privilege platform accounts, avoid reusable host passwords where possible, confirm create/delete/import/export actions before running them, and make sure agent logs and platform logs redact passwords and tokens.
SKILL.md:38Credential-Bearing API Requests Permit Plaintext HTTP Transport
SKILL.md:317Passwords Are Embedded in Process Command-Line Arguments
Referenced artifact was not completely inspected
详细参数见：`references/api_victoriametrics.md`
Referenced artifact was not completely inspected
详细参数见：`references/api_victoriametrics.md`
Referenced artifact was not completely inspected
详细说明见：`references/api_snmp_metrics.md`
Referenced artifact was not completely inspected
详细说明见：`references/api_snmp_metrics.md`
Referenced artifact was not completely inspected
详细流程见：`references/add-os-monitor.md`
Referenced artifact was not completely inspected
详细流程见：`references/add-os-monitor.md`
Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
TOKEN=$(curl -sS -X POST "${ZHIHU_API_URL}/system/auth/login" \
-H "Content-Type: application/json" \
--data '{"username":"'"${ZHIHU_USER}"'","password":"'"${ZHIHU_PASSWORD}"'","captchaVerification":""}' \
| python3 -c 'import sys,json; print((json.load(sys.stdin).get("data") or {}).get("accessToken", ""))')
Referenced artifact was not completely inspected
- v1.0.0 (2026-05-28): 初始 ClawHub 注册版本，整理 `SKILL.md` 和 `references/`。
Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
API_URL="${ZHIHU_API_URL:-http://localhost:1024}"
TOKEN=$(curl -s -X POST "${API_URL}/admin-api/system/auth/login" \
-H "Content-Type: application/json" \
-d "{
\"username\": \"${ZHIHU_USER:-admin}\",
Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
2a. 通过模型列表查询 modelId
MODEL_INFO=$(curl -s --connect-timeout 10 -X POST "${API_URL}/admin-api/cqt/asset-model/page" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer ${TOKEN}" \
-d '{"pageNo":1,"pageSize":50}' | python3 -c "
Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
2b. 通过 modelId 获取完整模型信息（含 items）
curl -s --connect-timeout 10 -X POST "${API_URL}/admin-api/cqt/asset-model/get" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer ${TOKEN}" \
-d "{\"id\": ${MODEL_INFO}}" | python3 -c "
The skill’s title and descriptive instructions are presented only in Chinese, and the rest of the document continues in that locale. Under the policy provided, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly documented and justified.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
智护平台 API 使用登录接口获取访问令牌：
curl -sS -X POST "${ZHIHU_API_URL}/system/auth/login" \
-H "Content-Type: application/json" \
--data '{
"username": "'"${ZHIHU_USER}"'",
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
智护平台 API 使用登录接口获取访问令牌：
curl -sS -X POST "${ZHIHU_API_URL}/system/auth/login" \
-H "Content-Type: application/json" \
--data '{
"username": "'"${ZHIHU_USER}"'",
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
--data '{"username":"'"${ZHIHU_USER}"'","password":"'"${ZHIHU_PASSWORD}"'","captchaVerification":""}' \
| python3 -c 'import sys,json; print((json.load(sys.stdin).get("data") or {}).get("accessToken", ""))')
curl -sS -X POST "${ZHIHU_API_URL}/dashboard/alarms/summary" \
-H "Authorization: Bearer ${TOKEN}" \
-H "Content-Type: application/json" \
--data '{}'
该技能要求用户提供 SSH 用户名和密码，并将其发送到平台 API，但没有在用户交互层明确提示这是敏感凭据、将被如何使用、发送到哪里以及是否会被存储。这会降低用户对高敏感数据处理的知情权，并增加凭据误交、误传或被不当记录的风险。
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
API_URL="${ZHIHU_API_URL:-http://localhost:1024}"
TOKEN=$(curl -s -X POST "${API_URL}/admin-api/system/auth/login" \
-H "Content-Type: application/json" \
-d "{
\"username\": \"${ZHIHU_USER:-admin}\",
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
2a. 通过模型列表查询 modelId
MODEL_INFO=$(curl -s --connect-timeout 10 -X POST "${API_URL}/admin-api/cqt/asset-model/page" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer ${TOKEN}" \
-d '{"pageNo":1,"pageSize":50}' | python3 -c "
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
2b. 通过 modelId 获取完整模型信息（含 items）
curl -s --connect-timeout 10 -X POST "${API_URL}/admin-api/cqt/asset-model/get" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer ${TOKEN}" \
-d "{\"id\": ${MODEL_INFO}}" | python3 -c "
该步骤会将用户提供的 SSH 用户名和密码发送到平台 API 进行连通性测试，这是高敏感凭据的外部传输。虽然业务上可能需要，但文档未要求使用 HTTPS，也未明确说明凭据的存储、日志处理和最小化保留策略；再加上默认 API 地址是明文 HTTP，本上下文使风险明显上升。
使用用户提供的 IP、端口、用户名、密码测试 SSH 连通性：
curl -s -X POST "${API_URL}/admin-api/zhihu/snmp/testConnect" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer ${TOKEN}" \
-d "{
This markdown file describes a delete operation for asset types but provides no warning about data removal, reversibility, or need for confirmation. For markdown files, destructive behaviors that can affect user data should be accompanied by a warning so users understand the impact before invoking the skill or API.
The file includes an endpoint for deleting assets but does not warn users that the action may remove asset records and impact system data integrity. Under the markdown-specific warning rule, destructive operations should be clearly disclosed to users.
The import and export sections describe moving asset data into and out of the system, which can affect user data and potentially expose sensitive inventory information, but no warning is provided. Markdown skill documentation should disclose behaviors that may affect privacy or system data handling.
This markdown file documents potentially impactful behaviors such as deleting scan tasks and exporting discovered asset data, but it does not include any warning about deletion effects, data sensitivity, or operator caution. Under the markdown-specific warning criterion, behaviors affecting user data or system integrity should be disclosed to users.
The create dialing task section allows users to configure HTTP/TCP/PING monitoring tasks, including a concrete HTTP request target, but does not warn that created tasks may generate ongoing network traffic and repeated checks against the specified systems. That omission is relevant for markdown under SQP-2 because it can affect external systems and operational integrity.