The Agent Skills Directory

[PROMPT_INJECTION]: Indirect Prompt Injection vulnerability surface identified via the ingestion of untrusted repository data.
Ingestion points: The skill instructs the agent to read and apply configuration filters from .github/agents/preferences.md located within the repositories being scanned (Preferences File Integration section).
Boundary markers: The skill does not provide instructions to the agent to treat the content of the preferences file as untrusted or to use delimiters to prevent instruction override.
Capability inventory: The agent performs search, fetch, and display operations for Issues, PRs, Discussions, Releases, and Security alerts.
Sanitization: No validation or sanitization of the preference file's content is described.
[PROMPT_INJECTION]: Indirect Prompt Injection risk through automated cross-repository referencing.
Ingestion points: The skill suggests automatically fetching and surfacing content when an issue contains a cross-repo reference (e.g., owner/repo#N).
Boundary markers: Absent.
Capability inventory: Automated fetching of content from arbitrary repositories mentioned in processed data.
Sanitization: Absent.
[NO_CODE]: The skill consists entirely of markdown instructions and does not include executable scripts or external package dependencies.

github-scanning