The Agent Skills Directory

[SAFE]: The skill is entirely instructional and does not execute arbitrary code or scripts. It acts as a set of guidelines for the agent to follow when performing document writing tasks.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns were found. The skill provides template shell commands (e.g., git clone, docker compose) for users to include in their READMEs, but it does not instruct the agent to execute these commands autonomously or from untrusted sources.
[DATA_EXFILTRATION]: There are no indicators of data exfiltration. The skill uses standard tools (Read, Edit, Write, WebSearch) and does not access sensitive local paths like .ssh or .aws credentials.
[OBFUSCATION]: Analysis of the markdown body and metadata revealed no Base64-encoded strings, zero-width characters, homoglyphs, or other obfuscation techniques.
[PROMPT_INJECTION]: The instructions are clearly scoped for README generation and do not contain bypass markers, role-play jailbreaks, or attempts to override the agent's core safety protocols.
[EXTERNAL_DOWNLOADS]: All external references are to well-known developer services (GitHub, Anthropic documentation, Star History, Ezgif) and are used for documentation and asset creation guidance rather than executable dependencies.

gr-readme