The Agent Skills Directory

[SAFE]: The skill's primary function is as a routing shell for internal technical documentation. It provides guidance to an AI agent on how to write idiomatic Rust code following specific workspace conventions.
[SAFE]: No hardcoded credentials, API keys, or sensitive file paths (e.g., SSH keys, .env files) were found. While the reference files contain code examples for handling credentials (e.g., SCRAM authentication), these are purely instructional templates for building a database engine and do not contain real secrets.
[SAFE]: All identified dependencies are standard, well-maintained libraries from the Rust ecosystem (e.g., tokio, arrow, datafusion, tantivy, crossbeam, zerocopy). There are no signs of typosquatting or usage of risky third-party packages.
[SAFE]: The skill emphasizes safe Rust practices. The 'unsafe' block policy requires mandatory '// SAFETY:' comments and validation using the Miri interpreter, which is a significant security control in systems programming.
[SAFE]: No obfuscation techniques or suspicious encoded payloads were detected. Binary constants found in the code (e.g., magic bytes for file formats like 0x5345474D) are standard practice for binary file format identification.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it ingests untrusted code from a workspace (via the Read tool) and possesses exploitable capabilities like Bash and Write access. However, this is a standard risk factor for coding assistant tools and is mitigated by the skill's primary purpose of adhering to established architectural patterns. Ingestion points: Read and Grep tools on workspace files. Boundary markers: Absent. Capability inventory: Bash, Read, Write, Edit, Grep. Sanitization: Absent.

rust