The Agent Skills Directory

[SAFE]: The skill defines a structured workflow for requirement elicitation and specification drafting without any detected malicious patterns or dangerous command executions.
[INDIRECT_PROMPT_INJECTION]: The skill includes a surface for processing user-provided requirements into a .beyond-code/<slug>/spec.md file. It implements a strict scope constraint that forbids the agent from writing or modifying implementation code during this stage, which acts as a security control against malicious payloads in user input being translated into executable logic.
[COMMAND_EXECUTION]: The skill does not perform any shell command executions; its operations are limited to reading and writing Markdown specification files.

beyond-code-think