The Agent Skills Directory

[SAFE]: The skill uses established memory management tools (search_notes, read_note, write_note) to manage conversation context and does not request excessive permissions.
[SAFE]: No external network connections, unauthorized file system access patterns, or data exfiltration attempts were detected.
[SAFE]: The synthesis logic, which involves overwriting existing notes (overwrite=True), is consistent with the skill's stated purpose and does not exhibit malicious data destruction patterns.
[SAFE]: Analysis for obfuscation, including Base64 encoding, zero-width characters, and homoglyphs, returned no findings.
[SAFE]: The logic for deriving session IDs from transcript files is presented as an optional configuration for compatible hosts and does not involve harvesting sensitive system credentials.

memory-capture