ClawHub
Back to skill

Security audit

新用户产品诊断

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed product-audit workflow that uses browser actions, test accounts, logs, and disposable email in ways that fit its stated purpose.

Install only if you intend to run real browser-based product audits. Before running on production, get explicit authorization, use only synthetic/test identities, avoid payments, review the mail.tm disposable-email dependency, and delete or protect run logs, screenshots, verification emails, and generated credentials after the audit.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (9)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill advertises broad activation phrases such as “模拟新用户”, “产品动线诊断”, and “first-run audit”, which can overlap with ordinary requests for product evaluation or walkthrough help. This can cause unintended invocation of a high-impact skill that performs real browser actions, account registration, and data creation against live systems, increasing the chance of unnecessary external interaction or policy-bound operations.

Content

No source excerpt is available for this finding.

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
80% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · SKILL.md (reported line 54)May include surrounding context.

bash
RUN=~/audit-runs/$(date +%Y%m%d-%H%M)-zenstory-staging   # 末尾换成 <产品>-<环境>；放在任何仓库之外
mkdir -p "$RUN"/shots "$RUN"/logs && chmod 700 "$RUN"
printf 'RUN=%s\nPRODUCT=zenstory\nTARGET=staging\n' "$RUN" > "$RUN/env.sh"
echo "$RUN"                                              # 之后所有命令都用这个字面路径

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 54)May include surrounding context.

bash
RUN=~/audit-runs/$(date +%Y%m%d-%H%M)-zenstory-staging   # 末尾换成 <产品>-<环境>；放在任何仓库之外
mkdir -p "$RUN"/shots "$RUN"/logs && chmod 700 "$RUN"
printf 'RUN=%s\nPRODUCT=zenstory\nTARGET=staging\n' "$RUN" > "$RUN/env.sh"
echo "$RUN"                                              # 之后所有命令都用这个字面路径

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The document is written entirely in Chinese and later instructs evaluators to assess whether the email body language matches registration, but it does not offer any opt-in language preference or explain that the skill is intentionally limited to a Chinese-speaking context. This can violate language/locale policy when users or operators may expect language flexibility.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/disposable-inbox.md (reported line 54)May include surrounding context.

md
ADDR="audit-$PERSONA-$(date +%Y%m%d)-$RANDOM@$DOMAIN"
MAILPW=$(openssl rand -base64 18)
APPPW="Au$(openssl rand -hex 8)!9"                         # 满足常见密码规则
curl -fsS -X POST "$API/accounts" -H 'Content-Type: application/json' \
  -d "$(jq -n --arg a "$ADDR" --arg p "$MAILPW" '{address:$a,password:$p}')" >/dev/null
jq --arg p "$PERSONA" --arg a "$ADDR" --arg m "$MAILPW" --arg w "$APPPW" \
  '. + {($p): {email:$a, mail_password:$m, app_password:$w}}' "$CRED" > "$CRED.tmp"

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
80% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · references/disposable-inbox.md (reported line 58)May include surrounding context.

-d "$(jq -n --arg a "$ADDR" --arg p "$MAILPW" '{address:$a,password:$p}')" >/dev/null jq --arg p "$PERSONA" --arg a "$ADDR" --arg m "$MAILPW" --arg w "$APPPW"
'. + {($p): {email:$a, mail_password:$m, app_password:$w}}' "$CRED" > "$CRED.tmp" mv "$CRED.tmp" "$CRED" && chmod 600 "$CRED" echo "注册邮箱: $ADDR" # 只打印邮箱，不打印密码 SH

text

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The instructions explicitly save full verification email contents to a local file on disk. Even though the file is meant to be temporary, verification emails often contain one-time links, codes, and potentially identifying metadata; storing them in plaintext increases exposure through local compromise, backups, shell access, or accidental retention.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The entire rubric is written in Chinese and provides no indication that language choice is optional or that the skill is intentionally limited to a Chinese-speaking context. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The entire report template and writing requirements are written as mandatory Chinese output instructions, with no indication that users may choose another language. This is a natural-language locale policy concern because it imposes a specific language by default without opt-in or a documented region-specific justification.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.