No evidence of prompt injection, malware, or install-time execution was found. The static findings are false positives from Markdown code formatting, type-stub docstrings, SQL examples, and documented DQL fetch examples that use placeholder domains. The skill does describe network-capable DQL scripts, so users should approve real endpoints and data sharing before running generated scripts.
No evidence of prompt injection, malware, or install-time execution was found. The static findings are false positives from Markdown code formatting, type-stub docstrings, SQL examples, and documented DQL fetch examples that use placeholder domains. The skill does describe network-capable DQL scripts, so users should approve real endpoints and data sharing before running generated scripts.
Static external-command and weak-crypto alerts were evaluated as false positives from Markdown code fences, inline code, and type-stub documentation. The skill intentionally documents outbound HTTP through fetch(), including authenticated requests and examples that send query results to external APIs, so publication should include a data egress warning.
The skill can generate SQL queries through query(), which can be risky if users concatenate untrusted input. The reference explicitly documents parameterized query placeholders, reducing the concern to normal SQL-script review.
The query capability is real, but the documentation directly recommends parameterized SQL. No evidence found that the skill instructs users to bypass safe query construction.
Capability review items (3)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
The skill teaches DQL users to call external URLs with fetch(). One example combines query() results containing user identifiers and email addresses with a POST to an external API. This is legitimate for integrations, but generated scripts could expose internal data if users choose untrusted endpoints or include sensitive fields.
The documentation directly presents outbound HTTP as a core capability and includes a database-to-external-API sync example. This is not malicious by itself, but the data egress path is explicit.
The HTTP reference includes examples with an Authorization bearer token placeholder and a form body containing username and password values. These are example credentials, not real secrets, but users should avoid embedding real credentials in generated scripts.
The values appear to be placeholders in documentation, so confirmed harm is low. The pattern is still relevant because the skill helps generate scripts that may include headers and request bodies.
The static analyzer flagged many backticks as Ruby or shell execution. Review showed these are Markdown inline-code markers and fenced DQL examples, not executable shell commands in the skill package.
The cited text is documentation syntax and sample DQL, with no script file or runtime path that executes shell commands. No evidence found for subprocess, shell, or OS command execution.
Static false positives ignored (1)
These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.
The weak-cryptography alerts matched ordinary API documentation terms such as hash-like wording, statistical mode, and smoothing alpha. No evidence found for cryptographic code, digest generation, or use of weak algorithms.
The inspected locations are type-stub and data-analysis documentation, not cryptographic implementations. The scanner labels do not match semantic context.