The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill performs a version check by fetching a JSON file from the official Canner/WrenAI GitHub repository. This is a vendor-owned resource used for maintaining the skill's currency.
[COMMAND_EXECUTION]: The skill executes several wren CLI tools and standard shell commands (such as grep, mkdir, and test) to manage project state and validate configurations. These operations are appropriate for the skill's purpose and are confined to the project environment.
[DATA_EXFILTRATION]: The skill reads the local ~/.wren/config.yml file to determine the active project path. This information is used locally to scope the skill's operations and is not exfiltrated to external endpoints.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by ingesting untrusted data from the raw/ directory (Step 4 in SKILL.md) to generate project instructions and metadata.
Ingestion points: Document files (PDFs, glossaries, handbooks) located in the project's raw/ folder.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are used when reading these files.
Capability inventory: The skill can perform SQL queries, write to YAML configuration files, and execute build/validation commands.
Sanitization: The skill lacks explicit sanitization of the text content extracted from user documents before it influences the agent's logic.

wren-enrich-context