The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest data from various project files to generate durable memory records. This creates a surface where malicious instructions embedded in those source files could influence the agent's behavior during the recording or subsequent retrieval phases.
Ingestion points: Data is extracted from notes.md, runtime phase sheets, audit findings, and fresh discovery notes (SKILL.md).
Boundary markers: The skill does not provide instructions for using delimiters or explicit warnings to ignore embedded instructions within the source data.
Capability inventory: The skill uses the doctrine command-line tool to read and write records to the .doctrine/ directory and manage links between memories.
Sanitization: There are no documented steps for sanitizing, validating, or escaping the content extracted from external files before it is committed to memory.

record-memory