The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands via the Bash tool to perform initial setup, specifically using mkdir -p to create a nested directory structure in .claude/memory/ for data storage.
[PROMPT_INJECTION]: The instructions contain a directive to override default agent platform behavior by modifying .claude/settings.json to set "autoMemoryEnabled": false. This is a functional requirement to enable the skill's own memory management system without conflict.
[INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where the agent automatically reads external files from the repository at the start of every session, creating a surface for indirect prompt injection if those files are modified by untrusted contributors.
Ingestion points: The agent is instructed to read .claude/memory/MEMORY.md and specific detail files within the .claude/memory/ directory automatically upon session start.
Boundary markers: No specific boundary markers or instructions to ignore embedded commands within the stored memory files are provided.
Capability inventory: The skill allows the agent to use Read, Write, Edit, Glob, Grep, and Bash tools.
Sanitization: There is no evidence of sanitization or validation of the content read from the memory files before it is processed as context.

memory-manager