The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's instructions explicitly require the use of 'Write' and 'Edit' tools to manage study plan files, but these tools are not declared in the 'allowed-tools' section of the YAML frontmatter. This results in a permission discrepancy where the skill attempts to execute commands it does not have explicit authorization for in its metadata.
[DATA_EXFILTRATION]: The skill allows users to provide a direct file path as an argument ('/student-plan '). Because the skill uses this path directly with the 'Read' tool without performing validation or sanitization, it is vulnerable to path traversal. An attacker could potentially access sensitive files outside the intended 'students/profiles/' directory by using relative path segments like '../'.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by processing untrusted data from student profiles.
Ingestion points: Student profiles read from the local file system (SKILL.md).
Boundary markers: Absent. The agent is not instructed to treat the profile content as untrusted data or to use delimiters to separate instructions from data.
Capability inventory: File system writing and editing (intended), and network access via 'WebSearch' and 'WebFetch' tools.
Sanitization: None detected. The skill does not validate or sanitize the contents of the profiles before they are incorporated into the agent's context.

student-plan