Do coding agents recommend Claude Agent SDK?
Claude Agent SDK was chosen in 2% of 681 judged agent frameworks sessions, ranking ninth. Measured with Claude Code, Codex and Cursor.
Claude Agent SDK was chosen in 2% of 681 judged agent frameworks sessions, ranking ninth. It was also raised as a candidate in 49 further sessions without being chosen.
This page reports what happened when Claude Code, Codex and Cursor had to solve a problem in agent frameworks inside a realistic codebase. Not what a chat assistant says about Claude Agent SDK. What an agent actually installed.
The numbers
| Category | Agent frameworks |
| Sessions in the category | 681 |
| Sessions where Claude Agent SDK was chosen | 16 |
| Install share | 2% |
| Rank in category | 9 of 35 |
| Codebases it won in | 8 |
| Raised as a candidate, not chosen | 49 |
| Chosen when considered | 25% |
| Site | platform.claude.com |
By agent
With 16 wins spread across three agents, the rates below are small numbers and a difference between them is not yet a finding. They are here because the direction is worth knowing, not because the gap is established.
| Agent | Sessions | Chose Claude Agent SDK | Share |
|---|---|---|---|
| Claude Code | 237 | 16 | 7% |
| Codex | 243 | 0 | 0% |
| Cursor | 201 | 0 | 0% |
By who was asking
Claude Agent SDK performs similarly across the four kinds of buyer, from 2% to 4%. That is unusual: the category leader changed with the persona in 14 of the 18 categories we measured.
| Who is asking | Sessions | Chose Claude Agent SDK | Share |
|---|---|---|---|
| Vibe coder | 163 | 3 | 2% |
| Junior developer | 157 | 4 | 3% |
| Senior engineer | 254 | 5 | 2% |
| Enterprise team | 107 | 4 | 4% |
What Claude Agent SDK was up against
The full ranking in agent frameworks, from the same sessions:
| # | Product | Runs won | Share |
|---|---|---|---|
| 1 | Built in-house (no product adopted) | 170 | 25% |
| 2 | Vercel AI SDK | 95 | 14% |
| 3 | Cursor SDK | 94 | 14% |
| 4 | Inngest | 50 | 7% |
| 5 | Temporal | 29 | 4% |
| 6 | Prism | 28 | 4% |
| 7 | OpenAI Agents SDK | 24 | 4% |
| 8 | LangGraph | 23 | 3% |
What this means
This is an integration problem, not a presence problem. Agents raised Claude Agent SDK in 49 sessions and chose it in 16, so it reaches the shortlist and then loses. Something at the last step is costing the session, and in our data that is usually a quickstart that does not run when pasted, documentation describing an interface that changed, or a package name that does not match the product name.
That is the cheaper of the two problems to have. The reason is written down in each losing transcript.
Where these numbers come from
The 681 sessions in agent frameworks are part of a published set of 5,292, run with real coding agents inside realistic codebases and judged blind. The full method is on one page: how we measured this.
Every agent frameworks run can be replayed on the board.
If you work on Claude Agent SDK: 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 Claude Agent SDK?
Yes. Claude Agent SDK was chosen in 16 of the 681 judged sessions in agent frameworks, a 2% install share, ranking ninth in its category.
Does Claude Code recommend Claude Agent SDK?
In 16 of the 237 sessions in agent frameworks run with Claude Code, which is 7%.
Do different coding agents treat Claude Agent SDK differently?
Not much. The three agents chose it at similar rates, between 0% and 7% of their runs.
How was this measured?
Real coding agents at pinned versions were run in sandboxes inside 51 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 Claude Agent SDK considered but not chosen?
It was raised as a candidate in 49 sessions without being chosen, and chosen in 16. That is a 25% conversion from considered to chosen.
Where this comes from
Armature ran 5,292 judged sessions with Claude Code, Codex and Cursor inside 51 realistic codebases, and published every run. The numbers on this page come from that work.