add-task
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local shell script
create-folders.shlocated within the plugin's own installation directory (${CLAUDE_PLUGIN_ROOT}) to initialize the required directory structure. This is a standard initialization procedure for the skill's functionality. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user input and writes it directly into a markdown file that is intended for later processing by other agent roles (e.g., a business analyst).
- Ingestion points: User input provided as a task description or title (SKILL.md).
- Boundary markers: The input is placed under a specific markdown header
## Initial User Promptto separate it from metadata. - Capability inventory: The skill has file writing capabilities (
Write tool) to the.specs/directory and shell execution (bash) for directory setup. - Sanitization: There is no explicit sanitization or escaping of the user-provided prompt content before it is written to the file, which could allow embedded instructions to persist in the task draft.
Audit Metadata