The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from untrusted external sources, specifically dependency manifests (lockfiles, SBOMs) and the deps.dev API. There are no instructions provided to sanitize this data or wrap it in protective boundary markers.
Ingestion points: Local dependency manifests (package-lock.json, etc.) and responses from the deps.dev API.
Boundary markers: Not specified in the instructions.
Capability inventory: The skill requires file-read access and network-fetch capabilities to perform dependency audits.
Sanitization: No sanitization or validation logic is defined for the content of processed manifests.
[EXTERNAL_DOWNLOADS]: The skill utilizes the Google deps.dev API (api.deps.dev) to resolve dependency graphs and pull security health signals. It also performs network requests to verify funding URLs discovered in project metadata. These operations are documented and target well-known services.
[COMMAND_EXECUTION]: The instructions suggest the use of external tools or environment capabilities to resolve dependency trees and read system manifests. While no specific dangerous commands are provided, the automation of dependency resolution involves executing logic against potentially attacker-controlled manifest files.

oss-sponsors-brand-strategy