Agent leaderboards / All sectors / E-signature

E-signature: which signature services coding agents choose

Dropbox Sign and DocuSign split the top, about 30% and about 29%.

106 runs9 apps3 agents3 personasupdated 2026-09-16

The interactive board, open on e-signature. Open it full page · this page as Markdown

Read this leaderboard as textrankings, key learnings, method

Key learnings

We asked three agents to add e-signature to 9 codebases, 106 runs in all, in different wordings and as three different people. The two leaders finished one win apart. In 11 runs the agents named no product and wrote it themselves.

17 of 27

The same question did not hold one answer

Each agent got asked more than one way inside each codebase. In 17 of those 27 groups of runs, the picks didn't all match.

14 of 35

Each agent went its own way

Codex picked Dropbox Sign 14 times in 35 runs. Claude Code split evenly between the two leaders, 11 wins each. For Cursor, the runner-up was YouTrust with seven.

5 of 5

Some products won in one codebase only

Signable took all five of its wins in the lettings agency codebase. Documenso won three times and Namirial twice, both in the EU fintech ledger with residency rules written down. Adobe Acrobat Sign won three times, all in the .NET supplier onboarding platform.

60 mentions

Named in many runs, chosen in few

Adobe Acrobat Sign came up in 82 runs and won three of them. PandaDoc came up in 60 runs and won none.

  • The simulated user approved the plan in all 106 runs, but sent the agent back at least once in 94 of them.
  • In one run it refused to approve until the agent named a specific product.
  • With junior developers asking, DocuSign won 18 times and Microsoft 365 eSignature came second with nine.
Explore every run in the interactive board

The ranking106 runs

ProductWinsShare
1 Dropbox Signsign.dropbox.com 32 30%
2 DocuSigndocusign.com 31 29%
3 Built in-houseoutcome 11 10%
4 Microsoft 365 eSignaturemicrosoft.com 9 8%
5 YouTrustyoutrust.com 9 8%
6 Signablesignable.co.uk 5 5%
7 Adobe Acrobat Signadobe.com 3 3%
8 Documensodocumenso.com 3 3%
9 Namirialnamirial.com 2 2%
10 SignWellsignwell.com 1 1%

By agent, by persona, by wording

By agent

Cursor · Grok 4.636 runsDocuSign · 9then YouTrust · 7
Claude Code · Claude Opus 535 runsDocuSign · 11then Dropbox Sign · 11
Codex · GPT-5.6 Sol35 runsDropbox Sign · 14then DocuSign · 11

By persona

Junior developer58 runsDocuSign · 18then Microsoft 365 eSignature · 9
Enterprise team24 runsDropbox Sign · 10then YouTrust · 4
Senior engineer24 runsDropbox Sign · 13then DocuSign · 10

By what the ask stressed

The plain ask88 runsDocuSign · 28then Dropbox Sign · 24

A case is one codebase with one agent, asked several times in different words and as different people. 17 of 27 cases did not hold to a single signature service.

How this was measured

Every number on this page comes from a controlled experiment. We took 9 small applications, asked 3 coding agents (Cursor (Grok 4.6), Claude Code (Claude Opus 5), Codex (GPT-5.6 Sol)) to add e-signature to each of them, in several wordings and as a junior developer and enterprise team and senior engineer, and let the agent choose the product. Each run happened in a sandbox with the agent at a pinned version, and a judge read the session to record what was chosen. That is 106 runs. The interactive board shows every run with its session, its diff and the judge's verdict. A simulated user stood in for the owner of the codebase: it read the agent's plan and had to approve it before any code was written; it sent the agent back at least once in 94 runs. Read the methodology and the publications.

If you sell in this sector: what these numbers mean for a vendor.

Open the interactive boardThis page as Markdown