Security audit
Security checks for vulnerabilities and agentic risk
This skill does natal chart calculations and optional Mystilink Wiki lookups in a way that matches its stated astrology purpose, with privacy caveats around birth data.
Before installing, treat birth date, time, and location as sensitive personal data. Use fictional or minimized data when possible, avoid sending raw birth profiles to the optional Wiki API, and only write output files to paths you intend to create or overwrite.
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
.idea/
.vscode/
*.log
.env
.env.*
!*.env.example
Without declared permissions the skill's intent is opaque and cannot be validated.
The README instructs users to collect and process birth time, date, location, and optionally query a remote Wiki API, but it does not warn that this is sensitive personal data or explain the privacy implications of transmitting derived astrological context to a network service. In a skill centered on natal chart generation, these inputs are intrinsically identifying and sensitive, so the omission can lead to unsafe handling, unintended disclosure, or privacy non-compliance by downstream users or hosts.
The README describes collecting birth time and place, then optionally querying a remote Wiki API, but it does not clearly warn that astrology birth data can be sensitive personal information. Even though the shown example requests only contain chart terms, the documented workflow encourages handling detailed birth profiles in a network-enabled skill without explicit data minimization, consent, or privacy guidance, which raises a real privacy risk.
The manifest specifies default_locale: en, and the documentation later reinforces English as the default with a Chinese fallback. This is a natural-language locale policy concern because the skill appears to impose a language default without explicitly offering the user a choice or documenting user opt-in.
Line L75 says that if the Wiki locale is omitted, it defaults to en, and may fall back to zh-Hans. This imposes a language/locale behavior without describing any user opt-in or selection mechanism, which matches the language/locale policy concern.
The README explicitly instructs use of an external Wiki API and gives example requests containing user-derived query terms, but it does not warn that those terms may be transmitted off-host. In this skill context, birth-chart interpretation can involve sensitive personal data or derived traits, so undocumented network disclosure creates a real privacy risk even if the examples themselves are harmless.
Documenting a silent fallback to English (and possibly zh-Hans) without user choice can cause user content or derived interpretation context to be routed in a locale the user did not request. While this is primarily a privacy and UX issue rather than a direct compromise, it can increase unintended disclosure and reduce transparency about how remote content is fetched.
This markdown file includes a natural-language locale policy that forces a language fallback when locale is not provided. Under the policy rules, imposing a specific language or locale without explicit user opt-in is a reportable issue unless it is clearly justified or optional.
The README explicitly documents optional outbound Wiki API requests but does not warn that user-derived astrology terms or profile context may be transmitted to an external service. In this skill, birth time and place are highly sensitive personal data, so omitting privacy and data-handling guidance can lead operators to unintentionally expose personal information or metadata to the network.
The reference shows API usage with locale=en on both documented endpoints, which implies the skill operates in English by default without mentioning user opt-in or a justified locale restriction. This can conflict with language/locale policy guidance when the file provides natural-language instructions that prescribe a single locale.
This code performs a file write when --output-file is supplied, but the implementation itself provides no confirmation prompt or explicit warning at the write site. For code-file SQP-2 checks, file writes should have some visible disclosure unless the operation is clearly communicated elsewhere in the skill description, which is not available in this file.
No suspicious patterns detected.