The Agent Skills Directory

[PROMPT_INJECTION]: The skill is designed to ingest and edit XML content from the Paligo API, creating a surface for indirect prompt injection.
Ingestion points: External data enters the agent context via the GET /documents/{id} endpoint described in SKILL.md, endpoints.md, and workflows.md.
Boundary markers: There are no explicit instructions or delimiters provided to the agent to prevent it from following potential instructions embedded within the ingested XML content.
Capability inventory: The skill includes Python scripts and recipes using the requests library in workflows.md to perform state-changing operations (PUT, POST, DELETE) on the API based on the data processed.
Sanitization: The skill does not implement content filtering or instruction-stripping on the XML content before the agent processes it.
[COMMAND_EXECUTION]: The skill instructs the user/agent to execute a locally provided Python script, scripts/validate_paligo_xml.py, for XML validation. This script uses the Python standard library and performs no dangerous network or system operations.
[SAFE]: All network communications are directed to the official domains of the service being integrated (paligoapp.com, paligo.net). No hardcoded credentials or unauthorized data exfiltration patterns were identified.

paligo-api