bmad-eval vs agent-evals
bmad-eval and agent-evals are both agent skills for LLM evaluation. bmad-eval is published by bmad-code-org and agent-evals by devops-skills. Both install in Claude Code, Claude apps, Codex and Cursor. On GitHub, bmad-eval has 54,505 stars and agent-evals has 48,428. Both were last active in Oct 2026. The table lines up where they differ (publisher, ships scripts and category) and has the install commands for both. Everything here comes from the listings' public sources and is refreshed daily. bmad-eval describes itself as: Runs the evals of a BMad skill through the agent harness the user works in and reports what they show: baseline against the bare model, variant against a…. agent-evals: Build automated evaluation suites for AI agents using golden datasets, rubrics, and regression gates. Licenses: bmad-eval unknown, agent-evals MIT.
Updated Oct 26, 2026 · Built from catalog data; how we rank and grade
bmad-eval
bmad-code-org
Runs the evals of a BMad skill through the agent harness the user works in and reports what they show: baseline against the bare model, variant against a…
agent-evals
devops-skills
Build automated evaluation suites for AI agents using golden datasets, rubrics, and regression gates.
Who should pick which
- bmad-eval and agent-evals cover the same ground; the table shows where they differ.
From listing data only. We don't pick a winner on security beyond the grades; read how grades work.
Side by side
| Fact | bmad-eval | agent-evals |
|---|---|---|
| Type | Skill | Skill |
| Publisher | bmad-code-org | devops-skills |
| License | License unknown | MIT |
| Pricing | free | free |
| Runs | Not stated | Not stated |
| Auth | Not stated | Not stated |
| Hosts supported | Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf | Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf |
| Tools | Not listed yet | Not listed yet |
| Security grade | –Security: not graded yet | –Security: not graded yet |
| Freshness grade | AFreshness: A | AFreshness: A |
| GitHub stars | 54,505 | 48,428 |
| Weekly downloads | Not measured | Not measured |
| Last release | Oct 14, 2026 | Oct 24, 2026 |
| Last activity | Oct 14, 2026 | Oct 24, 2026 |
Facts from each listing's public sources, checked daily. Tool lists arrive with tool scans.
Install
bmad-eval
npx skills add bmad-code-org/BMAD-METHOD --skill bmad-eval
Asks which agents to install it for (add -g for all projects). Needs Node.js 22.20+.
agent-evals
npx skills add sickn33/agentic-awesome-skills --skill agent-evals
Asks which agents to install it for (add -g for all projects). Needs Node.js 22.20+.
Also consider
- 0.784score
eval-improve
malloydata
Make the smallest safe edit that closes a diagnosed gap the model repo owns: a Malloy model edit, or a package skill under skills/.
SkillAFreshness: A★ 118Claude Code +9 - 0.784score
eval-answer
malloydata
Score one analytical answer against a verified golden, and score which of the entities the golden depends on retrieval delivered to the answerer.
SkillAFreshness: A★ 118Claude Code +9 - 0.784score
eval-diagnose
malloydata
Diagnose why a scored answer failed and who owns the fix, then cluster the failures by shared root cause.
SkillAFreshness: A★ 118Claude Code +9
Questions
- Is bmad-eval or agent-evals more popular?
- bmad-eval has more GitHub stars (54,505 against 48,428). Stars measure attention, not quality, so check freshness and docs too.
- Do bmad-eval and agent-evals work with Claude Code?
- Yes, both list a Claude Code install; the commands are on this page.
- Are bmad-eval and agent-evals open source?
- agent-evals is open source (MIT); the other is unknown.
Related
Cite this page
bmad-eval vs agent-evals. Agent Mart. https://agentmart.dev/compare/bmad-code-org-bmad-eval-vs-sickn33-agent-evals (last updated 2026-10-26)