The Agent Skills Directory

[SAFE]: The skill contains only documentation and architectural guidance for interacting with external MCP tools. No executable code, scripts, or binaries are present within the skill files.- [NO_CODE]: The skill consists of Markdown documentation and configuration metadata, with no associated executable logic or scripts.- [DATA_EXFILTRATION]: The documentation mentions standard environment variables like GITHUB_TOKEN for GitHub API rate limiting and GLUESTACK_PATH for local repository access. These are documented as standard configuration steps and do not indicate credential harvesting or unauthorized data transmission.- [PROMPT_INJECTION]: The skill describes tools that ingest external source code and metadata from repositories. This represents a surface for indirect prompt injection. 1. Ingestion points: Content is retrieved from external GitHub or local sources via 'get_component' and 'get_component_metadata' tools. 2. Boundary markers: No delimiters or ignore instructions are specified for the retrieved content. 3. Capability inventory: The skill's frontmatter specifies access to Bash, Write, and Edit tools. 4. Sanitization: No sanitization of the retrieved content is described in the provided workflows.

gluestack-mcp-tools