The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on executing a specialized suite of CLI tools (e.g., pythia similar, pythia src, pythia args, pythia cols, pythia unistr) to retrieve information from the codebase and database schema. These tools are used to facilitate the primary coding tasks and are consistent with the skill's intended purpose.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data by reading from existing source files and local configuration files, which could theoretically contain malicious instructions embedded as comments.
Ingestion points: Reads codebase source content via pythia src and project-specific house rules from .pythia/conventions.md.
Boundary markers: The instructions do not define specific delimiters or "ignore embedded instructions" warnings for the ingested codebase content.
Capability inventory: The skill has the capability to write files, interact with database metadata through tool calls, and generate PL/SQL code for deployment.
Sanitization: The skill explicitly mandates the use of bind variables for dynamic SQL to prevent injection attacks and provides a dedicated tool (pythia unistr) to handle string literals safely across different character sets.

pythia-write