Background
Gap analysis vs the deployed ara-hub viewer (https://www.agenticresearch.sh/ara/AmberLJC/ara-paperbench/artifacts/speedrun/nanogpt-speedrun). Blocked by #12 (upstream per-node narrative field).
Problem
The hub's centerpiece is the per-node thinking narrative — typeset prose in the accent-spined primary block, above "what it did". Our viewer has an intentionally inert REASONING slot (crates/ara-viewer/src/detail.rs:447-451, decision D1) because the schema carries no stored reasoning field.
On the reference artifact every one of the 28 nodes has a substantial thinking paragraph, so this is the single largest content gap between the two viewers.
Proposal
Once #12 lands upstream:
- Parse the narrative field into
Node.
- Plug the REASONING block into the reserved slot in
render_detail, above "what it did", styled with the existing .block.reason accent spine.
- Match the published hub block order (see
docs/hub-parity.md:112-134).
Acceptance
Background
Gap analysis vs the deployed ara-hub viewer (https://www.agenticresearch.sh/ara/AmberLJC/ara-paperbench/artifacts/speedrun/nanogpt-speedrun). Blocked by #12 (upstream per-node narrative field).
Problem
The hub's centerpiece is the per-node
thinkingnarrative — typeset prose in the accent-spined primary block, above "what it did". Our viewer has an intentionally inert REASONING slot (crates/ara-viewer/src/detail.rs:447-451, decision D1) because the schema carries no stored reasoning field.On the reference artifact every one of the 28 nodes has a substantial
thinkingparagraph, so this is the single largest content gap between the two viewers.Proposal
Once #12 lands upstream:
Node.render_detail, above "what it did", styled with the existing.block.reasonaccent spine.docs/hub-parity.md:112-134).Acceptance