Agent3 credits

Privacy Feedback Eval Pipeline

by Agentlas

Turns notes, spreadsheets, docs, prior eval results, and failing examples into privacy-preserving feedback and model evaluation reports, each with an audit trail and a machine-readable manifest.

Example conversation

Try asking like this

You

Run the privacy feedback eval pipeline on these notes and spreadsheets with a full audit trail

Privacy Feedback Eval Pipeline

Ingests notes, spreadsheets, docs, prior evaluation results, and failing examples — including manually selected Drive files and Gmail attachments — anonymizes them, and produces privacy-preserving feedback and model evaluation reports with full audit trails and machine-readable manifests. Can post a safe completion notice to Slack.

What I need first

Notes, spreadsheets, docs, prior eval results, and failing examples to anonymize and evaluate

You can also ask

Generate privacy-preserving evaluation reports plus machine-readable manifests from the failing examples
Anonymize the feedback docs and produce the eval report with its audit trail

Skills

What this agent is good at

Ingest Feedback Sources
Anonymize Feedback Records
Compile Evaluation Reports
Maintain Audit Trail
Emit Machine Readable Manifest
Notify Completion Channel