markdown-memory
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill establishes a workflow where the agent reads from and maintains a directory of Markdown files (
memory/lessons/) and anINDEX.mdfile. This pattern involves processing untrusted or historical data which could contain embedded instructions. - Ingestion points: Instructions in
SKILL.mdspecify readingINDEX.mdat session start and opening lesson files inmemory/lessons/during operation. - Boundary markers: Absent. The skill does not instruct the agent to use delimiters or block-level warnings to distinguish data from instructions.
- Capability inventory: The skill itself does not define scripts or tools, but the pattern is intended for use in agents with file system and project access.
- Sanitization: Absent. There are no guidelines for validating the content of the memory files or filtering potential injections.
- [NO_CODE]: The skill consists exclusively of Markdown instructions and does not include any executable scripts, package manager files, or binary components.
Audit Metadata