The Agent Skills Directory

[SAFE]: The skill consists of documentation and illustrative code snippets intended for developer education. No active malicious scripts or automated execution patterns were detected.
[SAFE]: External references point exclusively to official engineering blogs and documentation from Anthropic and OpenAI, which are well-known and trusted sources.
[SAFE]: No sensitive credentials, private file paths, or unauthorized data exfiltration patterns were identified in the instructional content or code examples.
[SAFE]: The context management strategies described (such as summarization and trimming) are standard industry practices. While they involve processing conversation history, the skill provides architectural guidance rather than implementing an exploitable runtime vulnerability.

context-engineering