The Agent Skills Directory

[DYNAMIC_EXECUTION]: The shell script located at scripts/gh-helpers.sh contains a function ghx_issue_create that dynamically constructs a command string from user-supplied arguments (title, body, labels) and executes it using eval. This pattern is susceptible to command injection if the input arguments contain shell metacharacters.
[COMMAND_EXECUTION]: The skill relies heavily on executing shell commands via the GitHub CLI (gh). Several files, including SKILL.md and references/automation-patterns.md, define complex aliases and shell scripts that perform network operations, file system modifications, and repository state changes.
[PRIVILEGE_ESCALATION]: Documentation in examples/example-gh-workflow.md instructs the user to execute commands with administrative privileges (sudo) to install the GitHub CLI and manage system-level repository keys. While these are official instructions from a trusted service, they represent a privilege escalation vector.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content from external sources (GitHub issues, pull requests, and comments) using commands like gh issue view and gh pr list. This creates a vulnerability surface where malicious instructions embedded in public GitHub content could be interpreted by the agent.
Ingestion points: Untrusted data enters the agent context through gh issue view, gh issue list, gh pr view, and gh pr list commands.
Boundary markers: The provided examples and instructions do not include boundary markers or system instructions to sanitize or ignore potentially malicious content within the fetched data.
Capability inventory: The skill possesses high-privilege capabilities, including the ability to create issues, merge pull requests, and perform GraphQL mutations across the repository.
Sanitization: There is no evidence of input validation or sanitization of the data retrieved from the GitHub API before it is presented to the agent for processing.

github-cli