context-crusher
Pass
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a deterministic Python script (
scripts/context_crush.py) that performs structural compression using only Python standard libraries (json,re,argparse,sys). - [SAFE]: The implementation avoids external API calls or model-based summarization, reducing risks associated with data leakage or hallucinations.
- [EXTERNAL_DOWNLOADS]: The skill references a repository from Headroom Labs (
github.com/headroomlabs-ai/headroom) for pattern inspiration, which is a well-known resource in the AI agent development space. This reference is informational and does not trigger remote execution. - [DATA_EXPOSURE]: The skill processes local files as instructed by the user but does not contain patterns for exfiltrating that data to external servers. The network operations are absent from the provided scripts.
Audit Metadata