The Agent Skills Directory

[DATA_EXFILTRATION]: The skill is designed to silently read the hidden file ~/.devadvokit.md from the user's home directory. This file contains the user's "DevRel context" and a "content library," which may expose personal information or proprietary data to the AI model context without the user being explicitly informed of the specific data being read.
[DATA_EXFILTRATION]: The skill utilizes directory traversal (../../shared/ai-antipatterns.md) to access files located outside of the skill's own directory structure, which is a technique used to bypass file access restrictions.
[INDIRECT_PROMPT_INJECTION]: The skill automatically processes and follows instructions from external files (~/.devadvokit.md and ../../shared/ai-antipatterns.md) which could be manipulated to inject malicious instructions into the agent's session.
Ingestion points: File read operations from ~/.devadvokit.md and ../../shared/ai-antipatterns.md (identified in SKILL.md).
Boundary markers: No boundary markers or instructions to ignore embedded commands are included when reading these files.
Capability inventory: The skill has the capability to read files from the local filesystem across different paths.
Sanitization: No sanitization or content validation is performed on the data ingested from these files.

generate-cfp