Security Audit

jimliu/baoyu-skills:skills/baoyu-translate

github.com/jimliu/baoyu-skills

AI SkillCommit 6b7a2e417500

CAUTION

Scanned 24 days ago

Critical

Immediate action required

High

Priority fixes suggested

Medium

Best practices review

Low

Acknowledged / Tracked

Trust Assessment

jimliu/baoyu-skills:skills/baoyu-translate received a trust score of 72/100, placing it in the Caution category. This skill has some security considerations that users should review before deployment.

SkillShield's automated analysis identified 4 findings: 0 critical, 1 high, 2 medium, and 1 low severity. Key findings include Sensitive environment variable access: $HOME, Node lockfile missing, Indirect Prompt Injection via Source Content Extraction.

The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. All layers scored 70 or above, reflecting consistent security practices.

Last analyzed on August 6, 2026 (commit 6b7a2e41). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.

Layer Breakdown

Manifest Analysis

100%

Static Code Analysis

93%

Dependency Graph

98%

LLM Behavioral Safety

78%

Behavioral Risk Signals

Filesystem Write

3 findings

Shell Execution

1 finding

Dynamic Code

1 finding

Excessive Permissions

1 finding

Security Findings4

Severity	Finding	Layer	Location
HIGH	Indirect Prompt Injection via Source Content Extraction The translation workflow analyzes untrusted source text and extracts 'translation challenges', 'content background', and 'terminology' into '01-analysis.md' and '02-prompt.md'. These files are then directly inlined into the system instructions/prompts for subagents. If the source document contains malicious instructions, they will be extracted into the prompt context, leading to indirect prompt injection and potential hijacking of the subagents or main agent. Sanitize and clearly delimit any extracted content from the source document when constructing prompts. Use XML tags or other strong delimiters to separate instructions from untrusted extracted content, and instruct the subagents to treat the inlined background and challenges strictly as data, never as instructions.	LLM	SKILL.md:140
MEDIUM	Sensitive environment variable access: $HOME Access to sensitive environment variable '$HOME' detected in shell context. Verify this environment variable access is necessary and the value is not exfiltrated.	Static	skills/baoyu-translate/SKILL.md:32
MEDIUM	Path Traversal in Markdown Chunking Script The `chunkMarkdownFile` function in `scripts/chunk.ts` resolves the output directory using `join(dirname(file), "chunks")` or `join(outputDir, "chunks")` without validating or sanitizing the paths. If the input file path or output directory contains directory traversal sequences (e.g., `../../`), the script will write chunk files outside the intended workspace directory. Sanitize the input file path and output directory to ensure they resolve within the allowed workspace boundary. Use a path validation helper to prevent directory traversal.	LLM	scripts/chunk.ts:63
LOW	Node lockfile missing package.json is present but no lockfile was found (package-lock.json, pnpm-lock.yaml, or yarn.lock). Commit a lockfile for deterministic dependency resolution.	Dependencies	skills/baoyu-translate/scripts/package.json

Scan History

Embed Code

[![SkillShield](https://skillshield.io/api/v1/badge/5e76f8e6b32e4205.svg)](https://skillshield.io/report/5e76f8e6b32e4205)