The Agent Skills Directory

[SAFE]: The skill functions as a reference guide for Go development. An analysis of the instructions, code snippets, and metadata reveals no evidence of prompt injection, data exfiltration, obfuscation, or persistence mechanisms.
[INDIRECT_PROMPT_INJECTION]: The skill documents patterns for ingesting and processing data from external sources (e.g., files, HTTP requests, and readers). While this defines a potential attack surface for indirect prompt injection if an agent uses these patterns to process untrusted data without further sanitization, the skill itself is purely instructional and does not introduce malicious behavior.
Ingestion points: ProcessData(r io.Reader), FetchWithTimeout(ctx, url), and LoadConfig(path string) snippets in SKILL.md demonstrate how to read data from external interfaces.
Boundary markers: Not applicable for these educational code snippets.
Capability inventory: The examples utilize standard Go library capabilities including os.ReadFile, http.DefaultClient.Do, and io.ReadAll for file and network operations.
Sanitization: The provided snippets are minimal idiomatic examples and do not include specific input validation or sanitization logic, which is standard for high-level pattern documentation.

golang-patterns