Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add kmgowda/SBK --skill sbk-distributed-benchmark-runnergit clone --depth 1 https://github.com/kmgowda/SBKWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/kmgowda/sbk/sbk-distributed-benchmark-runner)<a href="https://agentmods.dev/skills/kmgowda/sbk/sbk-distributed-benchmark-runner"><img src="https://agentmods.dev/badge/skills/kmgowda/sbk/sbk-distributed-benchmark-runner/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/kmgowda/sbk/sbk-distributed-benchmark-runner"><img src="https://agentmods.dev/badge/skills/kmgowda/sbk/sbk-distributed-benchmark-runner.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00105 | $0.01094 |
| Opus 5.5 | $0.00042 | $0.00438 |
| Sonnet 5.5 | $0.00021 | $0.00219 |
| Haiku 4.5 | $0.00011 | $0.00109 |
Grade A, and why
sbk-distributed-benchmark-runner scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
SBK Distributed Benchmark Runner
Follow the repository-wide safety and verification rules in AGENTS.md.
If diagnosis leads to a proposed code change, read
docs/HOT_PATHS.md. Remote SBK and SBM contain
the measurement hot paths; SBK-GEM itself is lifecycle/orchestration critical.
Obtain the exact explicit approval required by that inventory before adding
work to H0, H1, or H2 paths.
Use this skill for multiple load-generator hosts. For one load generator,
including one that connects to remote storage, use $sbk-benchmark-runner.
Select the distributed mode
- Use
sbk-gemwhen one controller should connect over SSH, reconcile Java and SBK, launch the same workload on every node, and aggregate through its embedded SBM. - Use
sbk-gem-yalfor the same orchestration from a reusable YML file. - Use standalone
sbmwhen another system launches SBK clients manually. Each client must selectGrpcLoggerand point to the SBM host/port. - Do not start a separate SBM for a normal SBK-GEM run; GEM owns one.
Read distributed-topology.md before any run. Use gem-workflows.md for GEM/GEM-YAL and standalone-sbm.md for manual aggregation. Use distributed-validation.md to preflight and judge the result.
For distributed Dell ECS/ObjectScale S3 load, also read
drivers/minio/docs/ECS_OBS_AGENT_RUNBOOK.md and
drivers/minio/docs/ECS_OBS_BENCHMARK_RUNBOOK.md
completely. Qualify the same ordinary MinIO-driver command on every load node
before adding GEM orchestration.
Required workflow
- Build the installed distribution with
./gradlew installDist. For standalone SBM also run./gradlew :sbm:installDist. - Make the intended SBK command succeed on one load node first.
- Confirm every load node can reach the backend.
- Confirm controller-to-node SSH, trusted host keys, remote disk/permissions, and Java 25.
- Confirm every node can reach the controller's advertised SBM host and port
(default
9717). - Start with one node, one worker, a dedicated target, and a short bounded run.
- Scale node and worker counts only after client connections, exact counts, return codes, and aggregate metrics are correct.
- Preserve sanitized topology, arguments/YML, versions, per-host responses, aggregate output, and backend/load-host telemetry.
What ships with it
6 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- today First seen · 84 lines · 105 tokens per session scan A 5cdef77f5339
sbk-distributed-benchmark-runner is a skill published in the GitHub repository kmgowda/SBK (36 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 105 tokens to every session and 1,094 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-11-07.
sbk-distributed-benchmark-runner compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| sbk-distributed-benchmark-runnerkmgowda/SBK | Skill | 36 | 1mo ago | Apache-2.0 | A | 105 | Codex |
| dotnet-reversezhaoxuya520/reverse-skill | Skill | 44k | 1mo ago | MIT | A | 144 | Claude Code |
| operating-github-ci-fixerTracer-Cloud/opensre | Skill | 13k | yesterday | Apache-2.0 | A | 48 | Claude Code, Codex |
| dart-run-static-analysisflutter/agent-plugins | Skill | 3.2k | 2d ago | BSD-3-Clause | A | 43 | Claude Code |
| agents-sdk-dotnet-debuggingmicrosoft/Agents | Skill | 1.1k | yesterday | MIT | A | 136 | Claude Code |
| memory-optimizationbenchflow-ai/skillsbench | Skill | 1.9k | 3d ago | Apache-2.0 | A | 60 | Claude Code, Codex |
| golang-error-handlingsamber/cc-skills-golang | Skill | 3.8k | 5d ago | MIT | A | 144 | Claude Code |