Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed development-automation workflow, but it needs review because it can drive code changes and commits through broad activation terms.
Review this before installing if you do not want a broadly triggered skill to guide agents through code changes, tests, and commits. It appears non-malicious, but users should invoke it only for deliberate plan execution and avoid putting secrets or sensitive internal details in plan text that will be copied into session tasks and shared with subagents.
The manifest text says the skill has 'removed risk code' and improved safety/stability, which implies a safer, de-risked capability set. However, the body of the skill directs subagents to implement code, run tests, commit changes, and relies on exec capability, all of which are side-effecting development actions rather than a contradiction-free 'risk removed' posture.
The trigger keywords are extremely broad development terms such as 'development', 'plans', and 'implementation', making accidental activation likely during ordinary coding conversations. In this skill, activation grants access to read and exec-guided workflow steps, increasing the chance that intrusive development automation or command execution is suggested in contexts where the user did not explicitly request it.
The workflow explicitly instructs the agent to read the plan, extract all tasks and context, and persist them in TodoWrite across the session. That creates a retention surface for potentially sensitive implementation details, requirements, or secrets embedded in plan files, and the same session then redistributes this context to multiple subagents, increasing exposure and propagation risk.
"Mark task complete in TodoWrite" [shape=box];
}
"Read plan, extract all tasks with full text, note context, create TodoWrite" [shape=box];
"More tasks remain?" [shape=diamond];
"Dispatch final code reviewer subagent for entire implementation" [shape=box];
"Use superpowers:finishing-a-development-branch" [shape=box style=filled fillcolor=lightgreen];
A substantial portion of the descriptive metadata is presented in Chinese, while the rest of the skill is in English, and there is no indication that the user can choose their preferred language. This can create a language policy issue if the skill effectively imposes a locale or language format without opt-in.
The manifest's differentiation statement says external dependency references were cleaned up, suggesting reduced reliance on outside components. Yet the skill materially depends on several other skills such as using-git-worktrees, writing-plans, requesting-code-review, and finishing-a-development-branch, so the documentation overstates independence from external references.
No suspicious patterns detected.