Trust Assessment
similarity-search-patterns received a trust score of 85/100, placing it in the Mostly Trusted category. This skill has passed most security checks with only minor considerations noted.
SkillShield's automated analysis identified 1 finding: 0 critical, 1 high, 0 medium, and 0 low severity. Key findings include SQL Injection in PgVectorStore.search due to unsanitized metadata key.
The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. All layers scored 70 or above, reflecting consistent security practices.
Last analyzed on February 11, 2026 (commit 5d65aa10). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings1
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| HIGH | SQL Injection in PgVectorStore.search due to unsanitized metadata key The `PgVectorStore.search` method constructs a SQL `WHERE` clause using an f-string to embed the `key` from `filter_metadata`. If `filter_metadata` is controlled by untrusted input, an attacker can inject arbitrary SQL into the query by manipulating the `key` value. This could lead to unauthorized data access, modification, or deletion within the PostgreSQL database. To prevent SQL injection, the `key` used in the `metadata->>'{key}'` expression must be properly sanitized or validated against a whitelist of allowed keys. Alternatively, use `asyncpg`'s `sql.Identifier` or `jsonb_extract_path_text` with parameterized path elements if dynamic keys are required, ensuring that the key itself is not directly interpolated into the query string. | Static | SKILL.md:210 |
Scan History
Embed Code
[](https://skillshield.io/report/8218cdc9889d760d)
Powered by SkillShield