Back to skill

Security audit

OpenClaw Daily Briefing

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward morning briefing helper, with some privacy and transport-security caveats users should understand before enabling optional integrations or scheduling.

Install this only if you are comfortable with a Chinese-language briefing workflow. If you enable the calendar, to-do, news, or scheduled briefing pieces, confirm which accounts are connected and whether sensitive meeting or task details will appear in the generated output. The weather helper should ideally use HTTPS and sanitize displayed responses before regular use.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/daily-briefing.sh:16
Finding

Plaintext Weather Request Allows Response Tampering and Terminal Output Injection

Content
View full analysis

Vulnerability Details

File Location: scripts/daily-briefing.sh, lines 16-18
Vulnerability Type: Plaintext external service communication and unsanitized terminal output
Risk Level: Medium

bash
# wttr.in 快速查询
WEATHER=$(curl -s wttr.in/?format="%t+%c+%h" 2>/dev/null || echo "暂不可用")
echo "- 当前:$WEATHER"

Technical Analysis

The scheme-less wttr.in URL causes curl to use plaintext HTTP. The request and response therefore lack transport encryption and server authentication. An attacker able to intercept the network connection can observe the request or replace the weather response.

The remote response is assigned directly to WEATHER and printed to the user's terminal without validation or sanitization. An attacker controlling the HTTP response can insert arbitrary text or terminal control sequences. Although the response is not evaluated as a shell command, crafted terminal sequences could alter terminal presentation, forge displayed messages, manipulate supported terminal features, or mislead the user.

Attack Path

  1. A user invokes scripts/daily-briefing.sh over an attacker-controlled or otherwise compromised network.
  2. The script connects to wttr.in using plaintext HTTP.
  3. A network-positioned attacker intercepts the request and supplies a forged response containing misleading content or terminal escape sequences.
  4. The forged response is stored in the WEATHER variable.
  5. The script prints the attacker-controlled content directly to the terminal.

Impact Assessment

Exploitation does not directly grant shell execution, elevated privileges, or filesystem access. The affected scope is the confidentiality and integrity of the weather request and the integrity of terminal output. An attacker can observe connection metadata, falsify briefing information, and potentially abuse terminal control sequences supported by the user's terminal emulator.

Remediation
View remediation

Remediation Suggestions

  • Require HTTPS explicitly and prohibit protocol downgrade:
    bash
    WEATHER=$(curl --fail --silent --show-error \
      --proto '=https' --tlsv1.2 \
      --connect-timeout 5 --max-time 10 \
      'https://wttr.in/?format=%t+%c+%h') || WEATHER='Temporarily unavailable'
    
  • Apply connection and total-operation timeouts so an unavailable service cannot stall scheduled executions.
  • Validate or sanitize the response before displaying it. Remove control characters and restrict output to the characters expected in a weather summary.
  • Keep the URL quoted to prevent shell metacharacter interpretation if query parameters are expanded later.
  • Log failures safely without exposing response bodies or sensitive environmental information.
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill description is written entirely in Chinese and the usage examples explicitly request Chinese output such as '生成今日晨间简报', with no indication that users may choose another language. This may violate a language/locale policy when the skill imposes a specific language without opt-in.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill aggregates multiple personal data sources—weather by location, calendar events, and to-do items—but the description does not warn users that sensitive personal and contextual information will be collected and combined into a single briefing. This increases privacy risk because aggregation can reveal routines, whereabouts, priorities, and meeting details more clearly than any one source alone, especially if the briefing is sent automatically on a schedule.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

Natural-language strings throughout the script, including status messages and prompts, are fixed to Chinese. This can violate language/locale policy when a skill forces a specific language without user opt-in or justification.

Content

No source excerpt is available for this finding.