The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from local files such as ./out/model.xlsx and ./templates/firm-template.pptx to generate presentation content. This represents an attack surface where maliciously crafted data in these files could influence agent behavior.
Ingestion points: The agent reads from ./out/model.xlsx and ./templates/firm-template.pptx.
Boundary markers: There are no instructions provided to separate or delimit untrusted data from instructions.
Capability inventory: The skill allows Python script execution via Bash and file writing to the local ./out/ directory.
Sanitization: The skill does not prescribe any validation or sanitization steps for the data read from external files.
[DYNAMIC_EXECUTION]: The skill instructs the agent to write a Python script and execute it using Bash at runtime. This involves dynamic script generation and execution to interface with the python-pptx library.
[COMMAND_EXECUTION]: The workflow involves using shell commands (Bash) to execute the generated Python scripts and perform directory management tasks like creating the ./out/ folder.

pptx-author