Security audit
Security checks for vulnerabilities and agentic risk
This is a testing guidance skill with broad activation phrases and some runnable examples, but its behavior is disclosed, purpose-aligned, and not hidden or destructive by design.
Install only if you want a broad QA/testing assistant. Review generated commands before running them, keep examples pointed at local or test environments, avoid real credentials or production data, and consider narrowing the trigger list if accidental activation would be disruptive.
The trigger 'test' directly overlaps with a built-in command namespace, creating a shadowing risk where user intent may be misrouted to this skill instead of the platform's expected command behavior. Such conflicts can lead to confusing or unsafe execution paths if the wrong capability is invoked.
The trigger list includes very generic terms such as 'test', 'testing', and 'QA', which are likely to appear in many unrelated user prompts. This can cause unintended activation of the skill, expanding its reach beyond the intended context and potentially interfering with safer or more specific skills or built-in behaviors.
The trigger 'test strategy' contains the reserved/built-in command term 'test' and may still collide with command parsing or prefix-based matching depending on the platform. This can cause ambiguous resolution between a built-in command and the custom skill.
The trigger 'test automation' begins with 'test', which may be interpreted as or collide with a built-in command if the system uses prefix or token-based matching. That ambiguity can lead to the wrong handler being selected for user requests.
The trigger 'test framework' similarly inherits the built-in command collision risk because it starts with 'test'. In systems with loose matching, users asking about framework work may inadvertently invoke this skill through a shadowed command path.
Additional triggers like 'manual testing', 'quality gate', 'defect', and 'regression' are broad operational terms that can match many normal engineering conversations. This increases the probability of accidental invocation and routing mistakes, especially in environments with multiple overlapping skills.
The trigger 'test maintenance' can conflict with built-in 'test' handling under prefix-based or partial matching schemes. This broadens the set of prompts that may be ambiguously routed and makes invocation behavior less predictable.
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.
Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.
**Test**: {Feature} in {language/locale}
- [ ] Text displays without truncation
- [ ] Date/time/currency formats correct
- [ ] Right-to-left layout (Arabic, Hebrew)
- [ ] Character encoding UTF-8
The trigger 'QA' is extremely short and ambiguous, making accidental matches likely in ordinary conversation, abbreviations, or unrelated text. While the direct impact is lower than command shadowing, it still increases unintended activation risk and reduces routing precision.
This markdown file contains a test setup example that posts generated user records to /api/test/seed, which is a network operation transmitting data. The surrounding documentation does not warn readers that the example sends data to an API endpoint or note that test seeding should be limited to non-production environments.
This markdown file includes an authentication example that posts credentials and reuses a bearer token in subsequent requests, but the surrounding documentation does not warn readers to use test accounts or non-production environments. For markdown files, safety-relevant behaviors affecting privacy or system integrity should be disclosed, and auth examples can encourage use against live systems without that caution.
Detected: suspicious.exposed_secret_literal