The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs shell commands for the gh CLI and curl using variables such as $REPO, $GHSA, and $NUM. These variables are extracted from local files like memory/vuln-scanned.json and files in memory/pending-disclosures/, which could lead to command injection if the underlying data files are maliciously modified.
[DATA_EXFILTRATION]: Details regarding private security advisories and repository vulnerabilities are transmitted to an external service using the ./notify script as part of the skill's monitoring functionality.
[PROMPT_INJECTION]: The skill ingests untrusted data from local files that could be used to influence the agent's behavior or notification style.
Ingestion points: memory/vuln-scanned.json, memory/pending-disclosures/*.md, soul/SOUL.md, and soul/STYLE.md.
Boundary markers: No explicit markers are used to separate untrusted data from processing instructions.
Capability inventory: Shell command execution (gh, curl), file system modification (git mv), and external notification dispatch (./notify).
Sanitization: No validation or sanitization of the ingested data is performed before it is used in command construction or notifications.
[EXTERNAL_DOWNLOADS]: Fetches repository metadata and security advisory status information from the official GitHub API via the gh CLI tool. These operations target a well-known service and are consistent with the skill's stated purpose.

vuln-tracker