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 ifBars/blender-agent-studio --skill blender-agent-benchmarkgit clone --depth 1 https://github.com/ifBars/blender-agent-studioWrote 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/ifbars/blender-agent-studio/blender-agent-benchmark)<a href="https://agentmods.dev/skills/ifbars/blender-agent-studio/blender-agent-benchmark"><img src="https://agentmods.dev/badge/skills/ifbars/blender-agent-studio/blender-agent-benchmark/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/ifbars/blender-agent-studio/blender-agent-benchmark"><img src="https://agentmods.dev/badge/skills/ifbars/blender-agent-studio/blender-agent-benchmark.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 | $0.00068 | $0.02630 |
| Opus 5.5 | $0.00027 | $0.01052 |
| Sonnet 5.5 | $0.00014 | $0.00526 |
| Haiku 4.5 | $0.00007 | $0.00263 |
Grade A, and why
blender-agent-benchmark 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 5d ago.
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 — 242 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Blender Agent Benchmark
Read the shared execution guidance once per task for autonomous decisions, evidence cadence, and long-task continuity.
Measure changes with the same tasks, model, effort, limits, Blender build, and evaluator. Preserve natural agent behavior.
The bundled runner selects the agent under test with --agent codex (default)
or --agent claude-code. Starting the runner from Claude Code does not change
that selection. Report the recorded agent CLI, requested model and observed
model identifiers; do not infer them from the host that launched the runner.
Protect benchmark integrity
- Create isolated directories for every condition and repetition.
- Do not leave the other condition's code, renders, metrics, or expected fixes where the agent can discover them.
- Keep the user-facing task prompt identical except for explicit skill invocation in the plugin condition.
- Use
codex exec --ignore-user-configfor the no-plugin baseline. - Use a pinned plugin directory in a fresh invocation for the plugin condition. The default is the plugin containing the runner;
--skill-rootselects a different snapshot. - Record CLI version, model, effort, Blender build, duration, tool calls, failures, and output hashes.
- Evaluate outputs after generation. Do not leak hidden rubric details to the agent.
Read references/methodology.md before changing fixtures, scoring, or comparison claims.
The opt-in scenes and game_characters suites add three tasks each, including
independent holdouts. Read scene and character protocol
for authored-camera evidence, render-only delivery and deformation checks.
Read references/open-source-benchmark-landscape.md
when designing new suites or borrowing evaluation ideas from other Blender
benchmarks.
Read references/validated-results.md only
when reviewing the plugin's recorded validation result, not while generating a
benchmark submission.
What ships with it
46 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.
- agents/openai.yaml 302 B
- references/astra-0.6-validation.md 4.1 KB
- references/astra-workflow.md 12 KB
- references/intake-cases.json 4.6 KB
- references/methodology.md 6.8 KB
- references/open-source-benchmark-landscape.md 7.8 KB
- references/quality-suites.md 2.6 KB
- references/scene-character-suites.md 2.6 KB
- references/validated-results.md 3.8 KB
- scripts/benchmark_mcp.ts 3.9 KB runs code
- scripts/claude-agent.test.ts 2.1 KB runs code
- scripts/claude-agent.ts 2.8 KB runs code
- scripts/compare_runs.test.ts 2.7 KB runs code
- scripts/compare_runs.ts 34 KB runs code
- scripts/evidence-provenance.test.ts 1.7 KB runs code
- scripts/evidence-provenance.ts 1.0 KB runs code
- scripts/inspect_game_scene.py 14 KB runs code
- scripts/model-options.ts 1.5 KB runs code
- scripts/motion-evidence.test.ts 4.2 KB runs code
- scripts/motion-evidence.ts 4.0 KB runs code
- scripts/pinned-mcp.test.ts 2.4 KB runs code
- scripts/pinned-mcp.ts 4.5 KB runs code
- scripts/provenance.test.ts 1.5 KB runs code
- scripts/provenance.ts 2.9 KB runs code
- scripts/quality-tasks.ts 14 KB runs code
- scripts/report_spatial_campaign.ts 3.2 KB runs code
- scripts/rescore_run.test.ts 2.3 KB runs code
- scripts/rescore_run.ts 6.5 KB runs code
- scripts/run_benchmark.test.ts 6.5 KB runs code
- scripts/run_benchmark.ts 30 KB runs code
- scripts/run_spatial_campaign.ts 4.0 KB runs code
- scripts/run_vanilla_baselines.ts 4.2 KB runs code
- scripts/run_vrchat_hair_benchmark.ts 6.3 KB runs code
- scripts/scene-character-tasks.ts 17 KB runs code
- scripts/scene-evidence.test.ts 5.4 KB runs code
- scripts/scene-evidence.ts 6.5 KB runs code
- scripts/scene-tasks.ts 9.7 KB runs code
- scripts/score.test.ts 11 KB runs code
- scripts/score.ts 19 KB runs code
- scripts/spatial-tasks.ts 5.8 KB runs code
- scripts/tasks.test.ts 4.7 KB runs code
- scripts/tasks.ts 54 KB runs code
- scripts/trace.test.ts 1.2 KB runs code
- scripts/trace.ts 2.8 KB runs code
- scripts/verified_score.test.ts 2.0 KB runs code
- scripts/verified_score.ts 3.1 KB runs code
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.
- 5d ago Changed ae838083f06b
- 20d ago Changed · +8 lines a2eb821f3143
- 21d ago Changed · +3 lines 2bf83ea10ea1
- 25d ago Changed · +4 lines 6f7b0bb50d4f
- 1mo ago Changed · +8 lines f2634ec6ab19
- 1mo ago Changed · +44 lines 8866f5d257ac
- 1mo ago Changed · +16 lines b136b4a63488
- 1mo ago First seen · 159 lines · 68 tokens per session scan A bc3909ccb51f
blender-agent-benchmark is a skill published in the GitHub repository ifBars/blender-agent-studio (94 stars, last pushed 2d ago), licensed MIT. It adds 68 tokens to every session and 2,630 once invoked, about $0.0003 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 2026-08-31.
Other skills, from other repositories
blender-validation
Blender scene, mesh, material, animation, and export readiness validation.
ai-prompt-regression-testing
Use when comparing prompt changes against a fixed evaluation set: record actual model settings, required checks, missing results, judge limitations and case-level regression evidence.
nemo-evaluator-sdk
Evaluates LLMs across 100+ benchmarks from 18+ harnesses (MMLU, HumanEval, GSM8K, safety, VLM) with multi-backend execution. Use when needing scalable evaluation on local Docker, Slurm HPC, or cloud platforms. NVIDIA's enterprise-grade platform with container-first architecture for reproducible benchmarking.
scenario-godot-expert
Use when an agent drives Godot 4.7 on macOS or Windows for any task: headless --script jobs, --import, --check-only, command-line exports, GUT tests, windowed screenshots, the headless editor (-e) or an MCP server into a running editor; when a run hangs or fails silently, a capture is black, Godot 3 code breaks in…
skills-validate
Use when explicitly asked to run a live or plan-only application test for a published Scenario skill.
validate
Validate a PartCAD part, assembly, scene or sketch against the requirements its requirements: section declares - run pc test for the ports, interfaces and physics PartCAD measures, run every simulation, render it from every angle that matters with its ports and interfaces drawn, read the bill of materials and the…