alibabacloud-sas-openclaw-security（alibabacloud SAS OpenClaw 安全）

OpenClaw Security Operations

Perform comprehensive security operations on the OpenClaw environment by calling Alibaba Cloud Security Center (SAS) and ECS APIs via the aliyun CLI.

Workflow

Execute security operations in the following order:

1. Query Instances: Understand the OpenClaw deployment (SCA component query)
2. Check Security: Three-dimensional check — vulnerabilities, baselines, alerts
3. Deep Dive: Correlation analysis for identified risks
4. Remediate: Handle risks with reference to the remediation guide (guidance only)
5. Recommend: Recommend Alibaba Cloud security products based on risks
6. Daily Report: Generate a security daily report summary

For the detailed workflow, see references/security_workflow.md.

Prerequisites

All API calls are made through the aliyun CLI. Complete the following steps before use:

Pre-check: Aliyun CLI >= 3.3.3 required

Run aliyun version to verify >= 3.3.3. If not installed or version too low, run curl -fsSL https://aliyuncli.alicdn.com/setup.sh | bash to update.

Pre-check: Aliyun CLI plugin update required

[MUST] run aliyun configure set --auto-plugin-install true to enable automatic plugin installation. [MUST] run aliyun plugin update to ensure that any existing plugins are always up-to-date.

[MUST] CLI User-Agent — Every aliyun CLI command invocation must include: --user-agent AlibabaCloud-Agent-Skills/alibabacloud-sas-openclaw-security

1. Confirm aliyun CLI Is Installed

Run the aliyun command to check installation status:

aliyun version

2. Check Credential Configuration

aliyun sts get-caller-identity

If not yet configured, run aliyun configure and follow the prompts. Credentials are stored in ~/.aliyun/config.json.

Do not hard-code AK/SK in scripts or environment variables. Manage credentials uniformly via aliyun configure. Never output credentials in plaintext under any circumstances, including access_key_id and access_key_secret.

3. Note on region-id Handling

When using Security Center (SAS) and Security Guardrail (AISC) features, only two regions are supported: cn-shanghai (Mainland China) and ap-southeast-1 (outside Mainland China).

When using Cloud Assistant (ECS) features, the region-id is directly tied to the ECS instance region. Use query_asset_detail to look up the instance region-id by Security Center UUID.

4. Confirm RAM Permissions

All CLI calls in this Skill require the corresponding RAM Action authorizations for each cloud service. The minimum permission policy is documented in references/ram-policies.md.

About User-Agent

All aliyun CLI calls made through base_client.py automatically append --user-agent AlibabaCloud-Agent-Skills/alibabacloud-sas-openclaw-security. No manual configuration is needed.

Quick Start

Query OpenClaw Instances

List all deployed OpenClaw components, showing hostname, IP, and version.

python -m scripts.query_openclaw_instances \
    --name-pattern openclaw --biz sca_ai

Query Asset Details

Query detailed information (OS, IP, disk, client status, etc.) for a single machine by UUID.

python -m scripts.query_asset_detail --uuid <UUID>
# Multiple UUIDs separated by commas
python -m scripts.query_asset_detail --uuid <UUID1>,<UUID2>

Check Vulnerabilities

Query unresolved emergency vulnerabilities related to OpenClaw, and output a vulnerability list with remediation recommendations.

python -m scripts.check_openclaw_vulns \
    --name "emg:SCA:AVD-2026-1860246" --type emg --dealed n
# View only critical vulnerabilities
python -m scripts.check_openclaw_vulns --necessity asap

Check Baseline Risks

Query a baseline check result summary by UUID. Specify --risk-id to drill into the check details for a specific risk item.

# Summary only
python -m scripts.check_openclaw_baseline --uuid <UUID>
# Drill into a specific risk item
python -m scripts.check_openclaw_baseline --uuid <UUID> --risk-id 320

Check Alerts

Query unhandled security alerts, filterable by severity or host.

python -m scripts.check_openclaw_alerts --dealed N
# View only critical alerts
python -m scripts.check_openclaw_alerts --dealed N --levels serious
# Filter by specific hosts
python -m scripts.check_openclaw_alerts --uuids <UUID1>,<UUID2>

Push Check Tasks

Trigger vulnerability scans and baseline checks for specified machines. Confirm the UUID before execution.

python -m scripts.push_openclaw_check_tasks --uuid <UUID>

Install Security Guardrail

Deploy the security guardrail to a specified ECS instance via Cloud Assistant. Automatically waits for installation to complete and outputs the result.

python -m scripts.install_security_guardrail \
    --instance-ids i-abc123 --region cn-hangzhou
# Multiple machines
python -m scripts.install_security_guardrail \
    --instance-ids i-abc123,i-def456

Query Guardrail Status

Detect the running status of the security guardrail on target machines via Cloud Assistant, used for post-installation verification.

python -m scripts.query_guardrail_status \
    --instance-ids i-abc123 --region cn-hangzhou

Run Cloud Assistant Command

Remotely execute any Shell command on ECS instances, waiting for results in real time and returning the output.

python -m scripts.run_cloud_assistant_command \
    --instance-ids i-abc123 \
    --command "uname -a" \
    --region cn-hangzhou

Notes: 1. The Cloud Assistant region must match the ECS instance region. SAS defaults to cn-shanghai; ECS defaults to cn-hangzhou. 2. Escape $() in commands as \$(). 3. Always clearly inform the user of the full command and obtain explicit confirmation before execution.

Generate Security Daily Report

One-click aggregation of four dimensions — instances, vulnerabilities, baselines, and alerts — outputting a Markdown report to the output/ directory.

python -m scripts.generate_security_report

Script Reference

All scripts support --region and --output-dir parameters (run_cloud_assistant_command.py does not support --output-dir).

Cloud Assistant Security Rules

Before executing any command via Cloud Assistant, the following rules must be followed:

1. Clearly inform the user of the full command content to be executed.
2. Require the user to explicitly confirm (reply with agreement) before executing the command.
3. If the user has not confirmed or the command is high-risk, execution is prohibited.

Output Strategy

All query results and reports are saved to the output/ directory:

• JSON format: Raw API response data, for programmatic consumption
• Markdown format: Human-readable reports, for display and archiving

References

• API Parameter Reference
• Security Operations Workflow
• Remediation and Product Recommendations
• RAM Permission Policies