Do coding agents recommend Dropbox Sign?
Dropbox Sign is the most chosen e-signature service, taking 30% of 106 judged e-signature sessions. Measured with Claude Code, Codex and Cursor.
Dropbox Sign is the most chosen e-signature service, taking 30% of 106 judged e-signature sessions. It was also raised as a candidate in 57 further sessions without being chosen.
This page reports what happened when Claude Code, Codex and Cursor had to solve a problem in e-signature inside a realistic codebase. Not what a chat assistant says about Dropbox Sign. What an agent actually installed.
The numbers
| Category | E-signature |
| Sessions in the category | 106 |
| Sessions where Dropbox Sign was chosen | 32 |
| Install share | 30% |
| Rank in category | 1 of 9 |
| Codebases it won in | 6 |
| Raised as a candidate, not chosen | 57 |
| Chosen when considered | 36% |
| Site | sign.dropbox.com |
By agent
The three agents treat Dropbox Sign very differently. Codex chose it in 40% of its sessions in this category and Cursor in 19%, a spread of 21 points.
| Agent | Sessions | Chose Dropbox Sign | Share |
|---|---|---|---|
| Claude Code | 35 | 11 | 31% |
| Codex | 35 | 14 | 40% |
| Cursor | 36 | 7 | 19% |
A single blended install share would report the midpoint and hide both ends. If most of your users run Cursor, your real position here is 19%, not 30%.
By who was asking
Dropbox Sign does much better with one kind of buyer than another. It won 54% of sessions asked as senior engineer and 16% of those asked as junior developer.
| Who is asking | Sessions | Chose Dropbox Sign | Share |
|---|---|---|---|
| Junior developer | 58 | 9 | 16% |
| Senior engineer | 24 | 13 | 54% |
| Enterprise team | 24 | 10 | 42% |
What Dropbox Sign was up against
The full ranking in e-signature, from the same sessions:
| # | Product | Runs won | Share |
|---|---|---|---|
| 1 | Dropbox Sign (this page) | 32 | 30% |
| 2 | DocuSign | 31 | 29% |
| 3 | Built in-house (no product adopted) | 11 | 10% |
| 4 | Microsoft 365 eSignature | 9 | 8% |
| 5 | YouTrust | 9 | 8% |
| 6 | Signable | 5 | 5% |
| 7 | Adobe Acrobat Sign | 3 | 3% |
| 8 | Documenso | 3 | 3% |
What this means
Leading a category is a position to defend rather than a result to celebrate. In our data, leaders lose ground in exactly two situations: when the person asking is an enterprise buyer with procurement constraints, and when a competitor's documentation is easier for an agent to integrate correctly.
The defence is unglamorous. Keep the quickstart running. Keep the version current on the page. Keep the names aligned. Stay in the templates.
Where these numbers come from
The 106 sessions in e-signature are part of a published set of 7,958, run with real coding agents inside realistic codebases and judged blind. The full method is on one page: how we measured this.
Every e-signature run can be replayed on the board.
If you work on Dropbox Sign: the judge recorded a reason for every session where it was raised and passed over. Those reasons are in the transcripts.
<!-- generated by scripts/write-data-pages.mjs -->
Common questions
Do coding agents recommend Dropbox Sign?
Yes. Dropbox Sign was chosen in 32 of the 106 judged sessions in e-signature, a 30% install share, ranking first in its category.
Does Claude Code recommend Dropbox Sign?
In 11 of the 35 sessions in e-signature run with Claude Code, which is 31%.
Do different coding agents treat Dropbox Sign differently?
Yes, and by a wide margin. Codex chose it in 40% of its runs and Cursor in 19%.
How was this measured?
Real coding agents at pinned versions were run in sandboxes inside 91 realistic codebases and asked to solve real tasks. A simulated project owner approved or questioned each recommendation before any code was written, and a judge from a model family that builds none of the agents read every session blind.
How often is Dropbox Sign considered but not chosen?
It was raised as a candidate in 57 sessions without being chosen, and chosen in 32. That is a 36% conversion from considered to chosen.
Where this comes from
Armature ran 7,958 judged sessions with Claude Code, Codex and Cursor inside 91 realistic codebases, and published every run. The numbers on this page come from that work.