📚 Agentic Paper

Run {{ run_id }} · initial state: {{ state }}

⚠️ Auto-simulation. Reviewers below are LLM agents; output is a self-review aid, not a substitute for human peer review.

{% if auto_mode_warnings %}

⚡

Auto-Mode Active

The selected profile needed providers you didn't supply keys for. Each affected tier was remapped to the closest equivalent on a provider you do have. The run is proceeding — costs and behaviour may differ from the original profile.

{{ w }}

{% endif %}

Live agent activity

One row per completed provider call. Tailing output_paper_review/{{ run_id }}/audit.jsonl via Server-Sent Events.

Anthropic agents stream their extended-thinking deltas live above (only works when the agent is routed to an Anthropic model with thinking enabled — see the routing per agent in the report).

Show report New review