The Agent Skills Directory

[SAFE]: Systematic review of all skill files, including SKILL.md, resources/methodology.md, resources/evaluators/rubric_grant_proposal.json, and resources/template.md, confirms the absence of executable scripts, command-line instructions, or network-enabled operations.
[SAFE]: The skill does not request access to any system tools or external APIs. Its functionality is restricted to providing structured text templates and evaluative feedback based on provided documentation.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze untrusted user data (grant proposals), constituting an ingestion point for potential prompt injections. However, the skill lacks any capabilities to execute commands, write to the filesystem, or perform network requests. The absence of an execution vector renders this attack surface safe.

grant-proposal-assistant