inquiry analyzer

AdvisoryAudited by Static analysis on May 10, 2026.

Overview

Detected: suspicious.dangerous_exec

Findings (2)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

ConcernHigh Confidence

ASI03: Identity and Privilege Abuse

What this means

The skill can read and operate inside the user’s logged-in Alibaba browser session, including customer inquiries and related account data.

Why it was flagged

The analyzer authenticates to the local OpenClaw browser bridge with a hard-coded token. Because the workflow requires the user to be logged into Alibaba, this grants the skill automation access to an authenticated business account without a clearly declared credential boundary.

Skill content

const RELAY_TOKEN = '856baea1afbe169e5eec0f6ecb5b90c77ddeb06b2abe1154'; ... const OPENCLAW_CDP_URL = 'http://127.0.0.1:18800'; ... headers: { 'x-openclaw-relay-token': RELAY_TOKEN }

Recommendation

Use only with a dedicated browser profile/account, rotate or externalize the relay token, and ensure the skill declares and documents this session access clearly.

ConcernHigh Confidence

ASI04: Agentic Supply Chain Vulnerabilities

What this means

The skill may fail or may use whatever playwright-core package is available in the local environment, which weakens reproducibility and dependency trust.

Why it was flagged

The code requires an external runtime package, but the provided install specification has no package install or lockfile and the SKILL.md claims the skill is self-contained with no external dependencies.

Skill content

const { chromium } = require('playwright-core');

Recommendation

Add a pinned package manifest/lockfile or document the exact trusted runtime dependency expected by OpenClaw.

NoteHigh Confidence

ASI02: Tool Misuse and Exploitation

What this means

Running the OKKI command may initiate background checks for multiple inquiries rather than only viewing existing data.

Why it was flagged

The OKKI workflow can automatically click the button to start a background investigation. This is aligned with the stated OKKI background-analysis purpose, but it is still an account action performed by browser automation.

Skill content

if (el.children.length === 0 && el.textContent.trim() === '发起背调') { el.click(); return true; }

Recommendation

Confirm the inquiry range and OKKI account implications before running, especially if background checks have cost, quota, or compliance impact.

NoteHigh Confidence

ASI05: Unexpected Code Execution

What this means

Installing and invoking the skill runs included JavaScript files on the local machine.

Why it was flagged

The entry script spawns a Node process to run the analyzer. This is expected for the skill’s local-script design and uses an argument array rather than a shell string, but it is still local code execution.

Skill content

const child = spawn('node', args, { stdio: 'inherit', cwd: workingDir });

Recommendation

Review the included scripts and run the skill only from a trusted package location.

NoteHigh Confidence

ASI06: Memory and Context Poisoning

What this means

Customer conversations and inquiry details may remain on disk after the analysis is complete.

Why it was flagged

The skill documents that it writes raw chat records to local files, along with reports and cache files. This is purpose-aligned, but it creates persistent local copies of sensitive inquiry/customer context.

Skill content

| 聊天记录 | chats/inquiry-*.txt | 原始聊天记录文本 |

Recommendation

Store outputs in a protected location, delete old reports/chats when no longer needed, and avoid running on shared machines.

Findings (2)

critical

suspicious.dangerous_exec

Location: scripts/run-analysis.js:98
Finding: Shell command execution detected (child_process).

critical

suspicious.dangerous_exec

Location: scripts/run-okki.js:96
Finding: Shell command execution detected (child_process).