The Agent Skills Directory

[COMMAND_EXECUTION]: The skill interacts with the local system and external services by executing standard development tools such as git, ripgrep, npm, python, go, dotnet, ruby, php, and platform-specific CLIs like gh and glab. The skill instructions explicitly mandate the use of argument arrays and forbid shell interpolation to prevent command injection vulnerabilities.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch metadata and security advisories from several well-known and trusted platforms, including GitHub, GitLab, npmjs.org, pypi.org, crates.io, proxy.golang.org, search.maven.org, api.nuget.org, rubygems.org, packagist.org, and api.osv.dev. These connections are necessary for the skill's primary function of researching software candidates.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external API responses and local repository files like manifests and lockfiles, which creates a potential surface for indirect prompt injection.
Ingestion points: User requests, API responses from package registries, and local repository files (manifests, lockfiles) identified in SKILL.md and references/adapters.json.
Boundary markers: The skill enforces a specific Markdown/YAML contract (find-reuse.request.v1) for its input.
Capability inventory: The skill has the ability to execute a wide range of CLI tools and make HTTP requests as defined in references/adapters.json.
Sanitization: The instructions in SKILL.md specifically require running commands as argument arrays and forbid shell interpolation, providing a safeguard against malicious payloads in external data.

find-reuse