The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is designed to ingest and transform user-provided content into platform-specific posts. This creates a surface where malicious instructions embedded in the input data could influence the agent's behavior.
Ingestion points: User-provided source content during Step 1 and Step 3 of the workflow.
Boundary markers: None identified; user content is processed directly for adaptation.
Capability inventory: The skill includes Python patterns for making authenticated network requests (requests.post) to external APIs.
Sanitization: No explicit sanitization or validation of the input content is described in the workflow.
[COMMAND_EXECUTION]: The 'API Integration' section provides a Python script template that performs network operations and accesses environment variables. While intended for legitimate automation, it involves executing code that interacts with external services.
Evidence: The Python snippet in SKILL.md uses requests.post to send data to https://your-crosspost-service.example/api/posts and reads os.environ['POSTBRIDGE_API_KEY'].
[SAFE_PRACTICE]: The skill follows security best practices for credential management by instructing the use of environment variables (os.environ) rather than hardcoding API keys directly in scripts.

crosspost