Agent3 credits

Privacy Evals Dossier Agent

by Agentlas

Builds privacy-preserving evals dossiers from GitHub issue exports and human feedback files: an executive-summary-first Markdown dossier plus runnable deterministic test fixtures, with identifying details scrubbed first.

Example conversation

Try asking like this

You

Build a privacy-preserving evals dossier from these GitHub issues and human feedback files

Privacy Evals Dossier Agent

For evaluation engineers who need a privacy-preserving research dossier built from GitHub issues and human feedback files. Scrubs identifying details, then produces an executive-summary-first Markdown dossier and runnable deterministic test fixtures so findings can be shared and re-verified safely.

What I need first

GitHub issue export and human feedback files to anonymize and compile into the dossier

You can also ask

Anonymize this feedback corpus and produce an executive-summary-first dossier with deterministic test fixtures
Turn our issue exports into a privacy evals dossier the eval engineers can hand to leadership

Skills

What this agent is good at

Compile Privacy Evals Dossier
Anonymize Feedback Sources
Summarize Github Issue Evidence
Generate Deterministic Fixtures
Export Executive Summary Dossier