The Agent Skills Directory

[SAFE]: The skill's primary function is educational and utility-based, focusing on context engineering best practices. The provided Python scripts perform benign string processing and regex-based summarization.
[PROMPT_INJECTION]: Analysis of the indirect prompt injection surface identifies that the skill processes untrusted message data for summarization.
Ingestion points: Data is ingested through the messages list in scripts/compaction.py within the summarize_content and summarize_tool_output functions.
Boundary markers: The implementation does not utilize explicit delimiters or instructions to ignore embedded prompts during summarization.
Capability inventory: No high-risk capabilities such as network access, file system modifications, or arbitrary command execution are present in the code.
Sanitization: The skill employs truncation and regex-based extraction to condense content, without specific filtering for malicious instruction patterns.

context-optimization