The Agent Skills Directory

[COMMAND_EXECUTION]: The script src/repomix-gh.sh contains instances where variables such as $REPO and $EXTRACT_DIR are used in shell commands (e.g., gh api, find, du) without sufficient quoting or sanitization. Specifically, the use of eval with the find command poses a risk of command injection if a maliciously crafted repository name or file path is processed.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch repository ZIP files from github.com and installation resources for the GitHub CLI from cli.github.com. These target well-known, trusted domains and are consistent with the skill's primary purpose.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8).
Ingestion points: Untrusted data is ingested via curl downloads of external GitHub repositories in src/repomix-gh.sh.
Boundary markers: The output file is generated without explicit boundary markers or warnings to the downstream LLM to ignore embedded instructions within the processed code.
Capability inventory: The skill has capabilities to execute bash, curl, gh, and repomix on the host system.
Sanitization: There is no evidence of content sanitization or filtering of the repository data before it is bundled into the final markdown document.

repomix-gh