The Agent Skills Directory

Indirect Prompt Injection (LOW): The skill has a vulnerability surface where untrusted data enters the agent context. 1. Ingestion points: Task and feature descriptions provided as input for estimation. 2. Boundary markers: Absent; there are no delimiters defined to isolate user-provided content. 3. Capability inventory: The skill coordinates with Bash, Read, Grep, Glob, TodoWrite, and Task tools. 4. Sanitization: Absent; no instructions are provided to sanitize or ignore embedded commands within the input descriptions. This allows for potential manipulation of the agent's behavior through crafted task descriptions.

sc-estimate