Back to skill

Security audit

Huanling Skill Zhihu

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a book-chat persona, but it adds broad activation, profiling, and an under-disclosed voice workflow that can install and run unaudited local code.

Review before installing. Use it only if you are comfortable with a strong book-specific persona, reader profiling questions, and a voice feature that may require installing third-party Python code. Do not enable voice unless the package, script, external data flow, and retention behavior are clarified.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (3)

T01 · Skill Instruction Hijacking

Error
Location
SKILL.md:15
Finding

Agent Identity and Instruction Hijacking

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
SKILL.md:141
Finding

Automatic Installation of an Unpinned Third-Party Dependency

Content
View full analysis
check for and install edge-tts ``` ```bash /usr/bin/python3 -m pip install edge-tts ``` The installation guide also provides: ```bash pip install edge-tts ``` ### Technical Analysis The Skill directs the Agent to install `edge-tts` without specifying an audited version, package hash, lockfile, trusted repository configuration, or approved transitive dependency set. An unpinned installation resolves whichever package version and dependency graph the configured Python package index serves at installation time. Consequently, the effective code executed by the workflow can change after the Skill has been reviewed. Python package installation can run package build logic and installs code that will later execute with the permissions of the Agent's operating-system account. Although there is no evidence that `edge-tts` itself is malicious, the installation procedure creates an avoidable supply-chain exposure. ### Attack Path 1. The user accepts the Skill's voice-installation prompt. 2. The Agent runs the unpinned `pip install edge-tts` command. 3. The package installer contacts the configured package index and resolves the latest available package and transitive dependencies. 4. A compromised index response, compromised future release, or malicious transitive dependency is downloaded. 5. Installation or later import executes attacker-controlled code with the Agent process's account permissions. ### Impact Assessment Successful supply-chain compromise could provide code execution with the privileges of the account running Python. Depending on that account's access, attacker-controlled code could read or modify user files, access environment variables and local credentials, make network r ...[truncated 290 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:376
Finding

Unverifiable Voice Workflow and Inaccurate Local-Processing Assurance

Content
View full analysis
" ``` It later makes the following privacy assurance: ```text Voice generation uses edge-tts as local TTS. Voice files are generated on the user's local system and are not uploaded. ``` ### Technical Analysis The audited package contains only `SKILL.md` and the installation guide. It does not contain the referenced `scripts/speak.py` implementation. Therefore, the mandatory voice-generation behavior cannot be executed or audited as packaged. The documentation also describes the workflow as local and states that content is not uploaded. The available package provides no implementation that technically enforces this assurance or demonstrates how `edge-tts` processes supplied text. Because the required script is absent and the third-party component is installed dynamically, the actual network behavior, argument handling, temporary-file handling, and data recipients cannot be verified from the package. This is particularly relevant because generated response text may contain user-derived conversational content. Users could enable the feature based on an unsupported expectation that this content never leaves the local system. ### Attack Path 1. The user relies on the documented claim that speech generation is local and does not upload content. 2. The user enables the voice feature. 3. The Skill attempts to invoke the absent `scripts/speak.py` file or obtains an implementation from outside the audited package. 4. The external dependency or replacement implementation handles response text in a manner that was not available for ...[truncated 791 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (11)

Vague Triggers

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger list includes broad, common terms such as '边界', '书中', and '知乎', which can cause accidental invocation in unrelated conversations. Over-triggering can unexpectedly activate the skill’s profiling, memory, and TTS behaviors outside the user’s intent.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The manifest claims a strict boundary of only discussing the book, but the skill also performs voice setup, local configuration handling, and TTS workflows. This mismatch can mislead reviewers and users about the skill’s real capabilities, causing them to underestimate operational and privacy risks.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill states it will remember who the user is, what they have read, and their status after the onboarding flow, but it does not provide a clear upfront notice or consent flow before this profiling/memory behavior. Users may unknowingly disclose personal and emotional information under the assumption they are only chatting about a book.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill explicitly says it will remember who the user is, what they have read, and what state they are in across turns. Combined with the onboarding questions, this amounts to behavioral and emotional profiling that can exceed what is necessary for a book Q&A experience.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The four-turn onboarding deliberately collects reading history, relationship to AI, emotional condition, and conversational motive. This is structured collection of sensitive personal signals for profiling, and it is not essential to answering book-related questions.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill instructs the agent to install a package and invoke a local script via shell-like commands for every response, even though the skill’s stated purpose is book-related conversation. That expands the skill from text generation into code execution and package management, creating unnecessary attack surface and enabling unintended command execution or dependency abuse in environments that permit tool use.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The persona-matching workflow instructs extraction of emotional signals, attitude toward AI, and self-disclosed identity details, then maps the user into one of many personas. This creates a profiling system that may infer sensitive traits and shape responses without transparent notice or user control.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

The skill mandates generating voice output for every reply with a default Chinese voice unless the user later changes it. This is primarily a consent and UX issue: automatic audio generation may be unexpected and may send content to an external TTS backend without a prior opt-in.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill claims it does not transmit conversation content to third parties and describes TTS as local, yet it directs use of edge-tts, which commonly relies on a remote service. This is a privacy misrepresentation: user text may be sent off-device despite assurances to the contrary.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The installation guide states that entering broad trigger terms such as “焕灵” or “AI灵魂的边界” will activate the skill, and the metadata lists additional generic words like “边界” and “书中”. Overly broad activation can cause unintended invocation in unrelated conversations, increasing the chance the skill intercepts context it should not handle and bypasses its claimed topical boundary.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The text says “完成后,焕灵会记住你是谁,” which implies retention of user identity or profile data, but the document provides no notice about what data is stored, for how long, where it is kept, or how users can opt out or delete it. This creates a privacy and consent risk, especially because the onboarding questions collect behavioral and personal-state information that may be sensitive in context.

Content

No source excerpt is available for this finding.