Skip to content

[1/3] fix(benchmarks): track docs pages agents read across all harnesses - #5330

Merged
benpankow merged 4 commits into
mainfrom
benchmarks/track-docs-reads
Oct 26, 2026
Merged

benpankow merged 4 commits into
mainfrom
benchmarks/track-docs-reads

Conversation

@benpankow

@benpankow benpankow commented Oct 23, 2026 •

Copy link
Copy Markdown
Collaborator

Summary

We want to know which shipped docs pages help coding agents in the authoring benchmarks, and which pages they never open. scripts/analyze.mjs already reported docs reads, but it parsed an older raw transcript shape. With the native Claude Code, Codex, and OpenCode harnesses, it finds zero docs reads. The script now reads tool calls from agent-eval's harness-neutral transcript.json. --docs ranks pages by how many runs opened them and lists shipped pages that no run opened.

These numbers cover benchmark tasks only, not real usage. The published results file does not change.

Validation

  • Ran the old and new analyze.mjs --docs on a fixture with one native-shaped transcript per harness. Each transcript reads README.md. Old: no pages. New: README.md in 3/3 runs, plus 108 unread shipped pages.

  • Added scripts/analyze.test.mjs. It builds transcript.json with agent-eval's own parseTranscript for each harness and checks the aggregate and the unread list.

  • pnpm test in apps/benchmarks: 38 tests pass.

  • pnpm typecheck in apps/benchmarks, plus oxfmt and oxlint on the changed files: pass.

  • Ran the benchmark on this PR with kimi-k3 (OpenCode), claude-sonnet-5 (Claude Code), and gpt-6-sol-high (Codex) on weather-tool and openapi-connection: 6/6 passed. analyze.mjs on the artifacts lists the pages each run opened. Every run entered through README.md, then read tools/* or connections/* for its case.

    Model (harness) Case Result Turns Tools Tokens in / out Docs read
    claude-sonnet-5 (Claude Code) weather-tool passed 5 7 589k / 3.5k README.md, tools/overview.mdx
    claude-sonnet-5 (Claude Code) openapi-connection passed 5 7 646k / 1.9k connections/openapi.mdx, connections/overview.mdx, README.md
    gpt-6-sol-high (Codex) weather-tool passed 4 16 271k / 5.5k tools/human-in-the-loop.md ×2, tools/overview.mdx ×2, connections/overview.mdx, README.md
    gpt-6-sol-high (Codex) openapi-connection passed 3 11 184k / 1.7k connections/openapi.mdx, connections/overview.mdx, README.md
    kimi-k3 (OpenCode) weather-tool passed 6 12 189k / 1.8k README.md, tools/human-in-the-loop.md, tools/overview.mdx
    kimi-k3 (OpenCode) openapi-connection passed 3 13 155k / 1.2k connections/openapi.mdx, connections/overview.mdx, README.md

    Tokens in include cache reads.

Checklist

  • I added tests and documentation where relevant
  • I added a changeset if this touches the published eve package
  • DCO sign-off passes for every commit (git commit --signoff)

@vercel

vercel Bot commented Oct 23, 2026 •

Copy link
Copy Markdown
Contributor

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated
eve-docs Ready Ready Preview, v0 Oct 26, 2026 10:06pm UTC
1 Skipped Deployment
Project Deployment Actions Updated
eve-pkg Skipped Skipped v0 Oct 26, 2026 10:06pm UTC

Comment thread apps/benchmarks/scripts/analyze.mjs Resolved
@benpankow
benpankow force-pushed the benchmarks/track-docs-reads branch from ddf6bb8 to 39e11a2 Compare October 26, 2026 20:25
@vercel
vercel Bot temporarily deployed to Preview – eve-pkg October 26, 2026 20:25 Inactive
@vercel
vercel Bot temporarily deployed to Preview – eve-pkg October 26, 2026 21:09 Inactive
@benpankow
benpankow added this pull request to stack #5382 October 26, 2026 21:16
@benpankow benpankow changed the title fix(benchmarks): track docs pages agents read across all harnesses [1/2] fix(benchmarks): track docs pages agents read across all harnesses Oct 26, 2026
@benpankow benpankow changed the title [1/2] fix(benchmarks): track docs pages agents read across all harnesses [1/3] fix(benchmarks): track docs pages agents read across all harnesses Oct 26, 2026
@benpankow
benpankow marked this pull request as ready for review October 26, 2026 21:33
@benpankow
benpankow requested a review from a team as a code owner October 26, 2026 21:33
@benpankow benpankow added the run-authoring-benchmark Run selected authoring benchmarks on this PR (authoring-benchmark.yml) label Oct 26, 2026
@benpankow
benpankow force-pushed the benchmarks/track-docs-reads branch from b543207 to 0d848c5 Compare October 26, 2026 21:41
@vercel
vercel Bot temporarily deployed to Preview – eve-pkg October 26, 2026 21:41 Inactive
@benpankow benpankow added run-authoring-benchmark Run selected authoring benchmarks on this PR (authoring-benchmark.yml) and removed run-authoring-benchmark Run selected authoring benchmarks on this PR (authoring-benchmark.yml) labels Oct 26, 2026
roboeve Bot and others added 4 commits October 26, 2026 15:00
Signed-off-by: benpankow <ben.pankow@vercel.com>
Signed-off-by: benpankow <ben.pankow@vercel.com>
Signed-off-by: benpankow <ben.pankow@vercel.com>
Signed-off-by: benpankow <ben.pankow@vercel.com>
@benpankow
benpankow force-pushed the benchmarks/track-docs-reads branch from 0d848c5 to fb5fc8d Compare October 26, 2026 22:00
@benpankow benpankow added run-authoring-benchmark Run selected authoring benchmarks on this PR (authoring-benchmark.yml) and removed run-authoring-benchmark Run selected authoring benchmarks on this PR (authoring-benchmark.yml) labels Oct 26, 2026
@vercel
vercel Bot temporarily deployed to Preview – eve-pkg October 26, 2026 22:03 Inactive
@benpankow
benpankow merged commit 32dab65 into main Oct 26, 2026
124 checks passed
@benpankow
benpankow deleted the benchmarks/track-docs-reads branch October 26, 2026 22:59

This branch was successfully deployed

1 active and 1 inactive deployments
Preview – eve-docs — fb5fc8de Deployed Oct 26, 2026 by vercel[bot]
Preview – eve-pkg — fb5fc8de Deployed Oct 26, 2026 by vercel[bot]
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

run-authoring-benchmark Run selected authoring benchmarks on this PR (authoring-benchmark.yml)

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants