The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill requires the agent to analyze between 10 and 15 project files to identify business-critical events for tracking. This creates a vulnerability surface where the agent could potentially encounter and follow malicious instructions embedded in the project files it reads. However, this risk is inherent to the skill's primary purpose of automated code integration and is mitigated by the structured workflow provided.
Ingestion points: The agent reads project files to create an event tracking plan as specified in references/basic-integration-1.0-begin.md.
Boundary markers: No explicit boundary markers are defined for the file analysis phase.
Capability inventory: The agent is granted permission to write and modify files to implement PostHog capture code in references/basic-integration-1.1-edit.md.
Sanitization: No explicit sanitization or validation of the ingested code content is mentioned.
[EXTERNAL_DOWNLOADS]: The skill documentation and example project recommend installing official PostHog packages and refer to official PostHog domains for data ingestion and project settings. These resources are provided by the skill's author and are standard for the intended integration.
Evidence: Installation of @posthog/react and posthog-node packages in references/EXAMPLE.md and references/tanstack-start.md.
[COMMAND_EXECUTION]: The integration process involves executing standard development commands, such as package installation and running local dev servers. The agent is also instructed to run linting or formatting scripts found in the project's package.json to verify its changes.
Evidence: Workflow steps involving npm install, npm run dev, and linting scripts in references/EXAMPLE.md and references/basic-integration-1.2-revise.md.

integration-tanstack-start