The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to read external documentation files (docs/*.md) to provide answers to user questions. This creates a vulnerability surface where instructions hidden within those documents could influence the agent's behavior.
Ingestion points: Documentation files are read as instructed in the 'Workflow' section of SKILL.md.
Boundary markers: The instructions lack specific delimiters or warnings to ignore instructions that might be embedded within the documentation content.
Capability inventory: The skill primarily performs file read operations and provides natural language responses; no evidence of network operations, file writes, or shell command execution was found in the analyzed files.
Sanitization: There is no mention of sanitizing, filtering, or validating the content read from the documentation files before it is processed by the AI.
[DATA_EXFILTRATION]: The skill uses extensive relative path traversal (../../../../docs/) in SKILL.md to access files located multiple levels above its own directory. While this is likely intended for navigating a specific project structure to find documentation, the pattern grants the agent access to files outside the skill's immediate scope.

super-help