The Agent Skills Directory

Command Execution: The skill utilizes a local Python utility (scripts/materialize_thesis_tracker.py) to transform structured JSON data into CSV and Excel formats. This script is used for artifact generation and operates on data internal to the skill's workflow.
Indirect Prompt Injection Surface: The skill is designed to ingest and process data from external sources such as regulatory filings, news articles, and research memos. This creates a potential attack surface where adversarial content in those documents could attempt to influence the agent's output.
Ingestion points: Data enters the system from external URLs (news/filings) and user-provided files as described in references/intake-and-source-priority.md.
Boundary markers: The workflow instructions in SKILL.md and references/workflow-core.md require explicit source labeling and separation of factual evidence from analyst judgment, which helps maintain analytical integrity.
Capability inventory: The skill has the capability to write to the local file system in the output directory via its report generation script.
Sanitization: Content is processed into structured formats (CSV/XLSX) using the csv module and openpyxl library, which provide standard escaping for these file types.
Dependency Management: The skill declares a requirement for openpyxl, a standard and well-known Python library for handling Excel files, used here to create formatted thesis tracker workbooks.

thesis-tracker