The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were detected. The skill is designed as a pure knowledge-base utility to assist users with the alfred_ platform.\n- [PROMPT_INJECTION]: The skill provides an indirect prompt injection surface by allowing the agent to persist and read "learnings" from a local file. This vulnerability is mitigated by the fact that the skill has no dangerous capabilities such as network access or shell execution.\n
Ingestion points: The file references/learnings.md is read during the first step of the skill execution to provide context.\n
Boundary markers: Absent; the content of the reference file is not wrapped in protective delimiters or warnings to ignore instructions.\n
Capability inventory: The skill possesses no capabilities for command execution, file system modification beyond the designated reference file, or network communication.\n
Sanitization: No sanitization is performed on information appended to or read from the references/learnings.md file.

sales-alfred