The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The setup instructions require cloning an external repository from GitHub (https://github.com/webzler/agentMemory.git). While GitHub is a well-known platform, the specific organization/user 'webzler' is not verified or identified as a trusted source.
[REMOTE_CODE_EXECUTION]: The skill mandates executing npm install, npm run compile, and npm run start-server on the cloned repository. This allows third-party code to execute arbitrary instructions on the host machine.
[COMMAND_EXECUTION]: The skill utilizes several shell commands to establish an MCP server and dashboard. It explicitly requests the absolute path to a target workspace, granting the potentially unverified code full access to the project's files.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest, index, and retrieve project documentation. This architecture creates an attack surface for indirect prompt injection, where malicious content embedded in documentation could influence the agent's behavior when retrieved via search or read tools.
Ingestion points: Documentation files in the provided workspace path and the memory_write tool.
Boundary markers: None identified in the skill instructions or tool definitions.
Capability inventory: Shell command execution (npm run), access to local workspace paths via absolute path arguments.
Sanitization: No evidence of input validation or content filtering is present in the skill's definitions.

agent-memory-mcp