The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the context field to execute git diff --stat HEAD~20. This is a local shell command execution used to gather information about recent repository changes for analysis.- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted data from git logs, commit messages, and TODO comments found within the codebase. An adversary could place malicious instructions in these locations to manipulate the agent's output. Mandatory Evidence Chain: 1. Ingestion points: Files in design/gdd/, production/sprints/, production/milestones/, git logs, and files containing TODO comments. 2. Boundary markers: Absent. 3. Capability inventory: Read, Glob, Grep tools and shell execution via the context field. 4. Sanitization: Absent.

scope-check