The Agent Skills Directory

[NO_CODE]: The skill consists entirely of Markdown instructions and documentation for an external MCP server. It contains no executable scripts, binaries, or automated configuration routines.
[PROMPT_INJECTION]: The skill provides an interface to ingest and process data from a user's workspace (via the remindb database), which constitutes an indirect prompt injection surface. This is a standard risk factor for retrieval-augmented generation tools.
Ingestion points: Tools such as MemoryFetch, MemorySearch, and MemoryFetchBatch are used to read content from the database and inject it into the agent's context.
Boundary markers: The instructions do not specify the use of delimiters (e.g., XML tags or unique tokens) or instructions to ignore embedded prompts within the retrieved content.
Capability inventory: The skill provides read-only access. Write operations are delegated to a separate memoize skill, which mitigates the risk of automated self-propagation or unauthorized data modification by this specific skill.
Sanitization: No evidence of input/output sanitization or validation of the retrieved content is present in the static instructions.
[SAFE]: The skill includes a reference to documentation on GitHub (github.com/radimsem/remindb). This link belongs to the skill's author, is purely informational, and does not involve remote code execution or unauthorized downloads.

remind