Security checks across static analysis, malware telemetry, and agentic risk
This skill does what it says, but it can create full local archives of conversations, tool outputs, and environment details without built-in redaction or warning controls.
Install only if you are comfortable creating local, replayable archives that may contain secrets, personal data, file paths, project names, and tool output. Use it on conversations you have reviewed, keep exports out of synced or shared folders unless intended, and manually redact credentials or sensitive content before RAG ingestion, training use, or sharing.
This skill exports AI agent conversations into structured Markdown or JSONL archives for persistence, replay, debugging, and knowledge management. Behavior: * Reads active conversation context only when explicitly invoked by the user * Saves exports locally to project/session directories * Does not transmit conversation data to external services * Does not collect telemetry * Does not modify system configuration * Does not execute privileged or destructive operations Supported export formats: * Markdown * JSONL * Replay/session traces * RAG-ready chunked archives Primary use cases: * AI workflow archiving * Agent debugging * Conversation replay * Knowledge base generation * Long-term memory persistence
No static analysis findings were reported for this release.
VirusTotal findings are pending for this skill version.
