The Agent Skills Directory

[DATA_EXFILTRATION]: The skill includes usage examples that encourage users to upload sensitive configuration files to GitHub Gists.
Evidence: In SKILL.md, the 'Backup Configuration' section provides the command gh gist create ~/.bashrc ~/.vimrc --desc "My dotfiles backup".
Risk: Shell configuration files like .bashrc frequently contain sensitive environment variables, aliases, or hardcoded credentials. Uploading these to a remote service, even as a 'secret' gist, increases the attack surface for credential theft.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection by reading and processing content from external, user-controlled sources (GitHub Gists).
Ingestion points: The agent can retrieve external content using gh gist view and gh gist list (documented in SKILL.md).
Boundary markers: The instructions do not define boundary markers or instruct the agent to ignore embedded instructions within gist content.
Capability inventory: The skill has the capability to perform state-changing operations including file deletion (gh gist delete), content modification (gh gist edit), and arbitrary API calls (gh api) as documented in SKILL.md.
Sanitization: There is no evidence of content sanitization or validation of the retrieved gist data before it is processed by the agent.
Risk: A maliciously crafted gist could contain hidden instructions that influence the agent to perform unauthorized deletions or modifications when the gist is viewed or processed.

gist-management