skill-issues
Warn
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill's instructions involve executing shell commands such as
grepandsedusing user-supplied identifiers (<id>) and project data. There is no mention of sanitizing these inputs, which could lead to command injection if the agent executes the shell snippets with maliciously crafted arguments. - [INDIRECT_PROMPT_INJECTION]: The skill processes and displays content from
.octo/ISSUES.md, which serves as a persistence layer for issue descriptions and notes. - Ingestion points: The agent reads from
.octo/ISSUES.mdand.octo/STATE.mdduring the list, add, resolve, and show operations. - Boundary markers: No delimiters or instructions are provided to the agent to treat the retrieved issue content as data only, leaving it open to following instructions embedded within the issues.
- Capability inventory: The skill has access to shell execution and file system write capabilities.
- Sanitization: Input provided for issue descriptions or resolution notes is not validated or sanitized before being stored and subsequently processed.
Audit Metadata