The Agent Skills Directory

[SAFE]: No security issues detected. The skill provides architectural guidance, mental models, and testing protocols for the Yaklang DSL.
[COMMAND_EXECUTION]: The skill documents standard commands to execute Yak scripts using the internal engine (yak <file> or go run common/yak/cmd/yak.go <file>). These are provided as usage instructions for the intended primary purpose of the skill and do not involve malicious parameter injection or privilege escalation.
[EXTERNAL_DOWNLOADS]: The skill references official project documentation at yaklang.github.io and public repository resources. These references are informative and originate from trusted or well-known project sources.
[DATA_EXPOSURE]: The file contains a reference to a local development path (/Users/v1ll4n/Projects/yaklang). This is documented as a template for users to navigate to their local source code for testing and does not expose sensitive system credentials or configuration.
[INDIRECT_PROMPT_INJECTION]: The skill describes an attack surface where Yaklang scripts process untrusted HTTP traffic (MITM and Fuzzer hooks).
Ingestion points: Hook functions such as hijackHTTPRequest and beforeRequest ingest external network packets (SKILL.md).
Boundary markers: The YAK_MAIN convention provides a logical boundary to prevent test code from executing in production contexts.
Capability inventory: Scripts have capabilities to forward, drop, or modify traffic and interact with local databases (db.*).
Sanitization: The skill recommends using error handling (~) and assertions for validation. As a documentation skill, it guides safe script construction.

yak