The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches documentation, OpenAPI specifications, and reference skill definitions from official vendor domains (docs.mem0.ai) and GitHub repositories (github.com/mem0ai).
[COMMAND_EXECUTION]: Performs dependency installation, runs unit tests, and executes application lifecycle commands (start, ready_probe, write_call) as part of the integration verification process.
[INDIRECT_PROMPT_INJECTION]: The skill processes testing logic and shell commands from local files (plan.md, product.json) which define the verification workflow.
Ingestion points: Reads plan.md, product.json, and goal.md from the .mem0-integration/ directory in the current workspace.
Boundary markers: Output from tests and application logs is redirected into dedicated log files under the .mem0-integration/ directory.
Capability inventory: The skill executes package managers (pip, npm), test runners (pytest, jest), Docker Compose, and arbitrary shell commands defined in the local integration plan.
Sanitization: No explicit sanitization or filtering is performed on command strings retrieved from the workspace artifacts before execution.
[DYNAMIC_EXECUTION]: Dynamically executes shell commands for application testing that are assembled from the local plan.md configuration file to drive the end-to-end verification flow.

mem0-test-integration