The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill performs an automated analysis of the user's source code to extract integration details such as agent IDs and variable mappings. This ingestion of untrusted data creates a surface for indirect prompt injection, where malicious instructions embedded in code comments or data strings could attempt to influence the agent's behavior during the workflow creation process.
Ingestion points: The skill reads local source code files in Step 1 to identify existing /call API implementations.
Boundary markers: No specific boundary markers or instructions are defined to separate code content from the agent's internal reasoning.
Capability inventory: The skill utilizes curl for network operations, enabling it to create, modify, and publish workflows on the Bolna platform, as well as initiate real phone calls.
Sanitization: The skill uses jq to ensure valid JSON structure when building payloads and includes a mandatory validation step using the vendor's /validate API endpoint to catch configuration errors before deployment.
[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using curl and jq. These commands are used to interact with the Bolna API for managing workflow revisions, validating definitions, and conducting test runs. These operations are within the expected scope of an API migration tool.
[EXTERNAL_DOWNLOADS]: The skill fetches agent configurations and workflow node definitions from api.bolna.ai. These requests target the official vendor infrastructure and are necessary for the migration workflow.

migrate-call-to-workflow