The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting and acting upon content from project memory files and source code, creating a surface for indirect prompt injection if those files contain adversarial instructions.
Ingestion points: read_memory, find_symbol, search_for_pattern, and other tools that read local file content into the agent's context.
Boundary markers: The skill lacks explicit boundary markers or instructions for the agent to disregard natural language commands embedded within the data it processes.
Capability inventory: The skill possesses significant modification capabilities, including code refactoring (replace_symbol_body, rename_symbol), memory modification (edit_memory, delete_memory), and external updates (notion-update-page).
Sanitization: No evidence of sanitization or validation of the ingested content before it is used to influence agent behavior.
[DATA_EXFILTRATION]: The skill utilizes a notion-update-page tool to synchronize project task progress with Notion. While Notion is a well-known service, the automated transmission of project metadata and implementation status to an external platform should be noted.
[COMMAND_EXECUTION]: The skill includes and references a PowerShell script (scripts/serena-memory-backup.ps1) designed to back up memory files. While functional for its stated purpose, it represents an executable script distributed as part of the skill package.

serena-usage