The Agent Skills Directory

[NO_CODE]: This skill consists entirely of instructional markdown files and configuration metadata. It contains no executable scripts (Python, JavaScript, etc.), binaries, or shell commands, which effectively eliminates risks related to code execution or subprocess exploitation.- [PROMPT_INJECTION]: The skill demonstrates a surface for indirect prompt injection due to its core function of ingesting data from the project repository. * Ingestion points: The agent reads the repository's CONTEXT.md, CONTEXT-MAP.md, and project source code files to align modeling terminology. * Boundary markers: The instructions lack explicit delimitation or directives to disregard potential instructions embedded within the ingested repository data. * Capability inventory: The skill is authorized to perform file-write operations, specifically creating or updating documentation files (CONTEXT.md and ADR files in docs/adr/). * Sanitization: No explicit validation or filtering logic is prescribed for the data read from external project files before it is processed by the agent.- [SAFE]: Comprehensive analysis of the provided markdown and metadata files confirms that the skill does not attempt to access sensitive system paths (e.g., credentials, SSH keys), perform any network operations, or hardcode secrets. Its operations are strictly limited to the intended domain modeling and documentation management tasks.

domain-modeling