webmcp/analysis

Ingram Sheets

A spreadsheet whose engine runs in your browser and publishes twelve tools to whichever agent is driving the tab. Spreadsheets for models, not models for spreadsheets.

Aggregate 37
Leverage 10
Execution 9
Impact 9
Creativity 9

Each criterion 1–10, equally weighted; aggregate is their sum. Ranking is the pipeline's consolidated output.

01 Links & metadata

Category
productivity / productivity
Origin
built for the challenge / built for the challenge
Access
Eligibility
LIKELY_ELIGIBLE / LIKELY_ELIGIBLE
Substitution
MAJOR_DELTA / MAJOR_DELTA
Demo liveness
alive

Origin, access model, eligibility, and substitution are reviewer diagnostics, not judging criteria. Authentication requirements are not penalized.

02 The two blind reviews

Two independent reviewers scored this project blind, from a sanitized evidence packet. Scores are shown separately so the reasoning stays inspectable. A withheld score means the reviewers differed by more than two points.

Reviewer A (round 1)

confidence 72%
Leverage
8 /10

WebMCP exposes the live in-memory workbook and dependency-aware operations with changed-cell results, shared cursor/focus, and concurrent human edits; generic UI driving would be materially less reliable.

Evidence cited
  • IronCalc runs in-browser with no server-side workbook copy.
  • Writes return recalculation ripple and changed cells.
  • Person and agent operate on the same editable workbook state.
Execution
7 /10

The packet and frame sheet show a coherent conversational-to-spreadsheet workflow and a visible grid artifact, but no submitted video or end-to-end tool invocation evidence limits claimed completeness.

Evidence cited
  • Demo is reported alive and public repository exists.
  • Frame sheet shows prompt, assistant response, loading state, and spreadsheet artifact.
  • Description specifies twelve workbook operations.
Impact
8 /10

Spreadsheet users and model builders have a concrete need for reliable formula-preserving collaboration; the product directly addresses transcription, dependency, and synchronization errors.

Evidence cited
  • Stable cell addresses and dependency graph are central to the stated workflow.
  • Human edits are incorporated into subsequent agent results.
Creativity
8 /10

Treating the spreadsheet as a shared agent-human computational workspace, rather than a generated export, is a fresh and well-articulated interaction model.

Evidence cited
  • Shared editable intermediate state is emphasized.
  • Agent focus and annotations are represented in the sheet.

Reviewer B (round 2)

confidence 80%
Leverage
8 /10

WebMCP gives an external agent precise access to live workbook state and dependency recalculation, with intermediate changed-cell results and shared focus; UI automation is substantially less reliable.

Evidence cited
  • About text explains IronCalc WASM holds the live workbook in memory with no remote copy.
  • Writes return recalculation ripple and agent/user edits share the document.
  • Frame sheet visibly shows spreadsheet/table states.
Execution
7 /10

The architecture and interaction contract are coherent, and spreadsheet UI evidence is visible, but there is no video or transcript to prove the central agent-edit loop end to end.

Evidence cited
  • Demo alive and public repository exists.
  • Description specifies twelve tools and detailed write/recalculation behavior.
  • Devpost preview visibly shows populated spreadsheet UI.
Impact
8 /10

Model builders and spreadsheet users benefit from reliable formula-aware collaboration without copying data into a separate interpreter.

Evidence cited
  • Specific problem: stable cell addresses, dependency graph, editable intermediate values.
  • Shared human/agent editing and attached user changes address real workflow friction.
Creativity
8 /10

The product treats the spreadsheet engine and its dependency graph as the agent's execution substrate, not merely a canvas controlled by clicks.

Evidence cited
  • Recalculation ripple is part of the tool contract.
  • Agent focus color and edit attachment create a distinct collaborative interaction.

03 Review highlights

Standouts across reviewers

  • Structured recalculation feedback and shared live workbook state.
  • Clear explanation of why live local workbook state matters.
  • Correction-at-each-edit tool contract.

Red flags

  • No submitted video; frame evidence appears to show artifact generation but not WebMCP calls.
  • No video/transcript; several advanced behaviors are only described.

04 Evidence

What each artifact proves is labeled on the artifact itself. A Devpost page capture is packaging evidence, not proof the product runs; video frames are evidence from the submitted demo, not live verification.

Devpost page capture for Ingram Sheets
EX-01 Devpost page capture · packaging evidence, not runtime proof · source

EX-V Submitted demo video

Contact sheets from the ?s video the team submitted. This is what reviewers were shown; it demonstrates the product in motion but is not independent verification. · watch the original

Contact sheet 1 from the Ingram Sheets demo video
EX-V1 Sheet 1 of 3 · reported video evidence
Contact sheet 2 from the Ingram Sheets demo video
EX-V2 Sheet 2 of 3 · reported video evidence
Contact sheet 3 from the Ingram Sheets demo video
EX-V3 Sheet 3 of 3 · reported video evidence