The Agent Skills Directory

[COMMAND_EXECUTION]: In src/lib.ts, the harvestRepository function uses execSync to run a git clone command. The repoUrl argument, supplied by the user via command-line arguments, is interpolated directly into the command string without any sanitization or validation. This allows an attacker to execute arbitrary shell commands by providing a malicious URL containing shell metacharacters (e.g., ;, &, |, or backticks).
[PROMPT_INJECTION]: The skill clones and extracts data from untrusted external repositories to generate documentation and knowledge bases, creating a risk of indirect prompt injection.
Ingestion points: External files cloned from Git repositories (src/lib.ts).
Boundary markers: None identified in the source code or skill instructions.
Capability inventory: Subprocess execution (execSync), directory reading (fs.readdirSync), and file system write (safeWriteFile).
Sanitization: No sanitization or filtering of external content is performed before indexing or processing.

knowledge-harvester