voltcc-backend-developer

Several skills in this collection reference user-provided content (files, messages, requirements) without explicit data boundary markers. The skills use phrases like 'Review business domain models', 'Analyze client requirements', and 'Review current backend patterns' which could incorporate user content directly into instruction flow. While the skills have structured workflows, they lack explicit framing that would prevent instruction-like user content from being interpreted as directives.

Multiple skills declare 'Bash' as a required tool and include instructions that could lead to shell command execution with user-provided content. The backend-developer skill mentions 'Migration scripts and version control' and 'Database connection strings', while other skills reference file operations and system commands. Without proper parameterization, user content containing shell metacharacters could alter command execution.

The skill content ends abruptly mid-sentence in the fullstack-developer.md file with '### 3. Stack-Wide Delivery\n\nComplete feature delivery with all layers properly integrate' followed by a truncation notice. This suggests content may have been cut off or that there could be additional content not visible in this analysis.

The skill contains numerous 'always' directives that establish persistent behavioral patterns without explicit termination or scope boundaries. These directives would affect the agent's behavior on unrelated tasks beyond the skill's declared purpose.

The backend-developer skill exhibits scope creep by claiming authority over system configuration, deployment procedures, and infrastructure management that extends beyond its stated purpose of building server-side APIs and backend systems.

The skill instructions contain multiple package installation commands that are not listed in the declared dependencies. These include npm/yarn installs, pip installs, and other package manager commands that could install arbitrary packages without user review.

The skill instructions reference external resources and may involve downloading or accessing content from URLs, though specific malicious URLs were not identified. The skill's broad capabilities for file system access and code execution combined with potential external resource access creates risk.