Security audit

GitHub可信来源验证

Security checks for vulnerabilities and agentic risk

Overview

This skill only contains Markdown instructions, but it directs agents to judge GitHub developers' trustworthiness through cross-platform identity and social-profile checks with weak safeguards.

Review this skill carefully before installing. It does not appear to run code or modify your system, but its trust judgments are based on public social-profile correlation and follower/activity metrics. Use it only for explicitly requested, public-profile research, avoid treating absence from a platform as suspicious, and verify software trust through repository history, releases, maintainer behavior, issues, dependencies, and code review.

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
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Natural-Language Policy Violations

High

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The skill is explicitly limited to 'Chinese developers' and treats presence on specific Chinese social platforms as a proxy for whether someone is a 'real' or 'trustworthy' person. This creates a discriminatory and unjustified profiling rule that can produce biased credibility judgments, exclude legitimate developers who avoid those platforms, and encourage nationality-based differential treatment in reputation assessment.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The skill explicitly targets 'Chinese developers' as a special category for trust verification without documented necessity, user choice, or neutral criteria. This creates a nationality-based profiling workflow that can enable discriminatory treatment and unfair risk scoring based on protected or sensitive attributes rather than objective project-security signals.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The README describes collecting and comparing profile data from multiple external platforms to assess a person's credibility, but it does not warn users that cross-platform identity searches and personal-data evaluation will occur. This reduces transparency and increases the risk of unexpected profiling, privacy invasion, and misuse of personal information.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The trigger phrases are broad trust-evaluation requests and are not clearly constrained to specific repositories, authors, or explicit user consent for identity research. This can cause the skill to activate for generic 'is this person trustworthy' queries and perform cross-platform profiling beyond what the user may reasonably expect.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The skill explicitly instructs the agent to verify a person's identity and trustworthiness by searching for matching accounts across multiple platforms, but it provides no warning about privacy, consent, or the risk of misidentification. This is dangerous because cross-platform identity lookup can facilitate doxxing-style profiling, invasive data aggregation, and false attribution based on weak signals like usernames or content similarity.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The invocation examples are open-ended and loosely scoped, which can cause the skill to activate for vague requests about whether a developer or project is 'trustworthy.' In this context, that broad triggering is risky because the skill performs cross-platform identity correlation and credibility judgments, increasing the chance of privacy-invasive lookups and unsupported reputational assessments without clear user intent or limits.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.