The Agent Skills Directory

[PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill is designed to ingest untrusted data (specifications and requirements) and generate structured implementation plans that include executable code and shell commands. * Ingestion points: The skill body indicates it processes a 'spec or requirements for a multi-step task' (SKILL.md). * Boundary markers: The skill does not define specific delimiters or warnings to prevent the AI from obeying instructions embedded within the provided specifications. * Capability inventory: While this skill only performs planning, the output it generates is intended for consumption by execution agents (software-forge:executing-plans) and includes shell commands (pytest, git) and Python code blocks. * Sanitization: No sanitization or validation of the input requirements is performed before they are interpolated into the implementation plan template.

writing-plans