The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs and executes shell commands via the Bash tool using user-supplied input. Specifically, the --since argument is interpolated directly into commands like git log --since=<point>. This pattern creates a potential command injection surface if the agent platform does not provide underlying sanitization for shell strings.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from various external and potentially untrusted sources within a repository to generate summaries.
Ingestion points: Data is collected from git log output and files located in quality_reports/passports/, quality_reports/plans/, and quality_reports/session_logs/.
Boundary markers: The instructions do not define boundary markers or delimiters to isolate untrusted content from the system prompt.
Capability inventory: The skill has access to broad capabilities including Read, Write, Glob, Grep, and Bash tools.
Sanitization: There is no explicit requirement for the agent to sanitize or escape the content gathered from the repository before incorporating it into the final handoff brief, which could allow malicious instructions in commit messages or report files to influence the agent's behavior during the summary phase.

coauthor-brief