The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it reads and processes data from external project files to determine planning logic and task management.
Ingestion points: The skill reads data from prd.json and prd_backlog.json to calculate task counts and complexity scores.
Boundary markers: There are no specified delimiters or instructions to ignore embedded malicious content within the project files.
Capability inventory: The skill uses file reading capabilities and instructs the agent to create new skill files at Level 3 and Level 4 complexity, which involves file writing operations.
Sanitization: There is no evidence of sanitization or validation of the content retrieved from the JSON files before it is used to influence the agent's behavior.
[COMMAND_EXECUTION]: The skill instructs the agent to dynamically generate and create new skill files (e.g., domain-specific skill files) when the project scale is Large or Enterprise. This involves the agent writing new executable content to the workspace based on the project context.
[EXTERNAL_DOWNLOADS]: The skill includes a reference link to an external methodology on GitHub (github.com/bmad-code-org/BMAD-METHOD). This is a documentation reference and does not involve automated script execution or package installation.

pm-organization-scale-adaptive