The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from the local codebase (affected modules, patterns, configs) and project documentation (READMEs, prior PRDs, ADRs) to generate context for new project plans. If these source files contain malicious instructions, they could influence the agent's behavior during the brainstorming and PRD generation process.\n
Ingestion points: SKILL.md and REFERENCE.md (Step 1) identify affected modules and documentation as primary context sources.\n
Boundary markers: The skill does not specify the use of delimiters or 'ignore' instructions to separate content retrieved from files from the agent's internal logic.\n
Capability inventory: REFERENCE.md (Step 10a) demonstrates the capability to execute shell commands via the GitHub CLI (gh) and perform file system writes.\n
Sanitization: There are no explicit instructions to sanitize or validate the content retrieved from the codebase before it is interpolated into prompts or PRD fields.\n- [COMMAND_EXECUTION]: The skill automates project management tasks by generating shell commands for the GitHub CLI tool (gh). These commands interpolate strings such as the feature name and implementation goals directly into the CLI arguments.

think