honey-memory
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests repository data to generate and maintain a persistent documentation file, creating a surface where malicious instructions in the source code could be transferred to the project memory.
- Ingestion points: The agent is instructed to read the repository (source code, build configurations, and existing documentation) to derive architecture, build commands, and decisions.
- Capability inventory: The skill performs file system write operations at the repository root, specifically creating or updating
PROJECT.md,CLAUDE.md, or backup files (FILE.original.md). - Boundary markers: The skill does not specify the use of delimiters, blocks, or warnings within the generated markdown to isolate content derived from untrusted source code.
- Sanitization: There are no instructions to sanitize, escape, or validate the content extracted from the repository before writing it to the persistent memory file.
Audit Metadata