The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exposes a functionality to preview compression on arbitrary text payloads via the /api/compression/preview endpoint, which creates a surface for indirect prompt injection if the agent processes untrusted data through this utility.
Ingestion points: The text field in the JSON payload sent to the POST /api/compression/preview endpoint in SKILL.md.
Boundary markers: The instructions do not specify the use of delimiters or specific instructions for the agent to ignore potential instructions embedded within the text targeted for compression.
Capability inventory: The skill is primarily restricted to network operations via curl to interact with a compression API.
Sanitization: There is no evidence of sanitization, filtering, or validation of the input text before it is processed by the compression engine.

omni-compression