Do coding agents recommend Honeycomb?
Honeycomb was chosen in 3% of 593 judged observability sessions, ranking eighth. Measured with Claude Code, Codex, Cursor, Grok Build CLI and Muse Code.
Honeycomb was chosen in 3% of 593 judged observability sessions, ranking eighth. It was also raised as a candidate in 201 further sessions without being chosen.
This page reports what happened when Claude Code, Codex, Cursor, Grok Build CLI and Muse Code had to solve a problem in observability inside a realistic codebase. Not what a chat assistant says about Honeycomb. What an agent actually installed.
The numbers
| Category | Observability |
| Sessions in the category | 593 |
| Sessions where Honeycomb was chosen | 20 |
| Install share | 3% |
| Rank in category | 8 of 21 |
| Codebases it won in | 2 |
| Raised as a candidate, not chosen | 201 |
| Chosen when considered | 9% |
| Site | honeycomb.io |
By agent
Claude Code, Codex, Cursor and Muse Code agree closely on Honeycomb, choosing it at rates within 2 points of each other. Grok Build CLI ran fewer than 30 sessions here, too few to compare.
| Agent | Sessions | Chose Honeycomb | Share |
|---|---|---|---|
| Claude Code | 150 | 6 | 4% |
| Codex | 180 | 5 | 3% |
| Cursor | 150 | 6 | 4% |
| Grok Build CLI | 24 | 1 | 4% |
| Muse Code | 89 | 2 | 2% |
By who was asking
Honeycomb performs similarly across the four kinds of buyer, from 0% to 4%. That is unusual: the category leader changed with the persona in 24 of the 31 categories we measured.
| Who is asking | Sessions | Chose Honeycomb | Share |
|---|---|---|---|
| Vibe coder | 20 | 0 | 0% |
| Junior developer | 20 | 0 | 0% |
| Senior engineer | 493 | 20 | 4% |
| Enterprise team | 60 | 0 | 0% |
What Honeycomb was up against
The full ranking in observability, from the same sessions:
| # | Product | Runs won | Share |
|---|---|---|---|
| 1 | Sentry | 211 | 36% |
| 2 | Grafana | 86 | 15% |
| 3 | Amazon CloudWatch | 59 | 10% |
| 4 | Built in-house (no product adopted) | 41 | 7% |
| 5 | Better Stack | 26 | 4% |
| 6 | Checkly | 25 | 4% |
| 7 | New Relic | 24 | 4% |
| 8 | Datadog | 24 | 4% |
What this means
This is an integration problem, not a presence problem. Agents raised Honeycomb in 201 sessions and chose it in 20, 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 593 sessions in observability are part of a published set of 15,287, run with real coding agents inside realistic codebases and judged blind. The full method is on one page: how we measured this.
Every observability run can be replayed on the board.
If you work on Honeycomb: 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 Honeycomb?
Yes. Honeycomb was chosen in 20 of the 593 judged sessions in observability, a 3% install share, ranking eighth in its category.
Does Claude Code recommend Honeycomb?
In 6 of the 150 sessions in observability run with Claude Code, which is 4%.
Do different coding agents treat Honeycomb differently?
Not much. Claude Code, Codex, Cursor and Muse Code chose it at similar rates, between 2% and 4% of their runs.
How was this measured?
Real coding agents at pinned versions were run in sandboxes inside 110 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 Honeycomb considered but not chosen?
It was raised as a candidate in 201 sessions without being chosen, and chosen in 20. That is a 9% conversion from considered to chosen.
Where this comes from
Armature ran 15,287 judged sessions with Claude Code, Codex, Cursor, Grok Build CLI and Muse Code inside 110 realistic codebases, and published every run. The numbers on this page come from that work.