Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed multi-agent code review workflow that reads repository diffs and writes review reports, with no evidence of hidden, destructive, credential, network, or persistence behavior.
Install this if you want a heavier code-review workflow that may inspect repository context, launch five reviewer agents, and add Markdown reports to your repo. Review the generated docs/five-star-reviewers files before committing them.
Skill manipulates agent memory, state, or stored context. Memory corruption can alter personality, override safety rules, or cause unpredictable behavior.
### Promote to p0 when
- the change is likely to produce incorrect behavior
- an edge case can silently break a key path
- a missing failure guard can corrupt state or produce a bad external result
## `star_architecture`
The skill description is broad enough to be auto-invoked for many generic review requests, which can cause the agent to trigger unexpectedly and perform substantial actions such as launching multiple sub-agents and inspecting repository state. In this skill, that risk is amplified because invocation can cascade into file-writing behavior and broad repository analysis without a narrowly scoped user confirmation step.
The skill directs the agent to create multiple markdown files inside the current repository, but does not require explicit user consent or a warning that running the skill will modify the working tree. This can silently alter a user's repo, create noisy diffs, interfere with CI or review workflows, and be especially problematic when the skill is auto-invoked from broad review-related prompts.