The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process support ticket data, which contains untrusted content from external users that could be used for injection attacks.
Ingestion points: Data enters the agent's context through Freshdesk API responses, specifically the subject, description, and conversation_body fields documented in SKILL.md and references/tickets-api.md.
Boundary markers: The instructions do not define specific delimiters or boundary markers for the agent to use when interpreting these external data fields within its own reasoning process.
Capability inventory: The skill includes instructions for network operations via curl and local file writing (tickets.json) in the common workflow scripts in SKILL.md.
Sanitization: The skill identifies this risk under best practices in SKILL.md, recommending that users 'Sanitize user input
Prevent injection in descriptions', though no specific implementation is provided for the agent's internal prompt handling.
[EXTERNAL_DOWNLOADS]: The skill references several external SDKs and libraries to facilitate integration.
The documentation in SKILL.md and references/sdk-examples.md suggests installing community-maintained SDKs from GitHub repositories (sjkingo/python-freshdesk, kumarharsh/node-freshdesk, snkrheads/freshdesk-ruby) and official package registries (PyPI, NPM, RubyGems).
These references are for established community projects supporting the well-known Freshdesk service.
[COMMAND_EXECUTION]: SKILL.md contains functional Bash script examples for data management tasks.
Scripts utilize standard system utilities such as curl for API interaction, jq for JSON processing, and paste for file manipulation.
Examples include exporting tickets to a local tickets.json file and importing contacts from a contacts.csv file.

freshdesk-api