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 archetypeai/agent-skills --skill atai-newton-fusion-modelgit clone --depth 1 https://github.com/archetypeai/agent-skillsWrote 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/archetypeai/agent-skills/atai-newton-fusion-model)<a href="https://agentmods.dev/skills/archetypeai/agent-skills/atai-newton-fusion-model"><img src="https://agentmods.dev/badge/skills/archetypeai/agent-skills/atai-newton-fusion-model/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/archetypeai/agent-skills/atai-newton-fusion-model"><img src="https://agentmods.dev/badge/skills/archetypeai/agent-skills/atai-newton-fusion-model.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 | $0.00223 | $0.05087 |
| Opus 5.5 | $0.00089 | $0.02035 |
| Sonnet 5.5 | $0.00045 | $0.01017 |
| Haiku 4.5 | $0.00022 | $0.00509 |
Grade A, and why
atai-newton-fusion-model scanned grade A with 1 finding 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 7d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -X POST $ATAI_API_ENDPOINT/api/files \ How it starts
The opening of the file, as written. The whole thing — 296 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Newton C 2.6 Fusion Model — Multimodal /query in One Hop
Single stateless POST to /query for text, image, or video reasoning against the C 2.6 fusion checkpoint. No Lens session, no batch pipeline, no SSE. One call → one response.
When to Apply
- User wants to describe / classify / extract structure from an image with one API call
- User wants the same for a short video in a single
/querycall - User wants stateless multimodal reasoning (no per-session warmup, no orphan cleanup)
- User wants JSON output from a multimodal input (the prompt is the schema)
- User is building a serverless / per-request handler where a streaming Lens session would be the wrong shape
Do not use this skill when:
- The workload is live, streaming video → text from a long-running camera feed — a session-based activity monitor is the right shape, not stateless
/query - The workload is a large multi-file batch job
- The task is time-series classification via embeddings (KNN / anomaly detection)
The Model Identifier
Newton::c2_6_8b_fp8_260424d7a55d5e
The full Newton:: prefix and the _fp8_ segment are required. Variants that omit either are rejected with 400 invalid_model_version:
c2_6_8b_fp8_260424d7a55d5e ❌ missing Newton:: prefix
Newton::c2_6_8b_260424d7a55d5e ❌ missing _fp8_
c2_6_8b_260424d7a55d5e ❌ both missing
This was probed end-to-end against https://platform.u1.archetypeai.app/api/lens/query on prod.
Endpoint
POST {ATAI_API_ENDPOINT}/api/lens/query
Authorization: Bearer <API_KEY>
Content-Type: application/json
Both ATAI_API_KEY and ATAI_API_ENDPOINT are required — there is no default endpoint, so a wrong-deployment mistake fails loudly at startup instead of silently at query time. Prod is https://platform.u1.archetypeai.app/api. The same endpoint serves all three modalities below — what changes is the request body.
Wire Shapes
Text + JSON
{
"query": "Classify the operational state. Respond with ONLY the JSON object.",
"instruction_prompt": "You output {\"state\": \"...\", \"confidence\": <0..1>}.",
"file_ids": [],
"model": "Newton::c2_6_8b_fp8_260424d7a55d5e",
"max_new_tokens": 300
}
What ships with it
13 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.
- references/_common.py 5.5 KB runs code
- references/.env.example 216 B
- references/csv_vs_txt_proof.py 4.9 KB runs code
- references/image_query.py 5.6 KB runs code
- references/requirements.txt 112 B
- references/sample_assets/1_pass_2_pass_3_pass_B.mp4 7622 KB
- references/sample_assets/assembly_after.png 376 KB
- references/sample_assets/assembly_before.png 394 KB
- references/sample_assets/README.md 1.1 KB
- references/sample_assets/wind-turbines.png 1911 KB
- references/text_query.py 7.7 KB runs code
- references/video_query.py 5.6 KB runs code
- tests/test_references.py 11 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.
- 7d ago Changed 436c5f46c677
- 1mo ago First seen · 296 lines · 223 tokens per session scan A 0d95081e6a05
atai-newton-fusion-model is a skill published in the GitHub repository archetypeai/agent-skills (7 stars, last pushed today), licensed Apache-2.0. It adds 223 tokens to every session and 5,087 once invoked, about $0.0009 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
tao-finetune-kumo-anomaly
Kumo-Anomaly — diffusion-based anomaly detection and fine-tuning for multivariate time series. Use when the user asks to "fine-tune Kumo-Anomaly", "run Kumo-Anomaly inference", "detect anomalies with diffusion", "time series anomaly detection", "finetune Kumo-Anomaly", "use performanomalyanalysiswithdiffusion"…
tao-finetune-kumo-forecast
Kumo-Forecast — transformer-based multivariate time series forecasting with DARR (context-enhanced kNN retrieval), interpretability, and fine-tuning. Use when the user asks to "forecast with Kumo-Forecast", "run forecasting inference", "use performforecasting", "DARR mode", "context-enhanced forecasting", "lag horizon…
nixtla-model-benchmarker
Generate benchmarking pipelines to compare forecasting models and summarize accuracy/speed trade-offs. Use when evaluating TimeGPT vs StatsForecast/MLForecast/NeuralForecast on a dataset. Trigger with "benchmark models", "compare TimeGPT vs StatsForecast", or "model selection".
nixtla-research-assistant
Research and summarize Nixtla ecosystem updates and time-series forecasting content from the web and GitHub. Use when gathering release notes, recent changes, or best-practice references. Trigger with "Nixtla updates", "what's new with TimeGPT", or "find time-series papers".
timegpt-pipeline-builder
Generate production-ready TimeGPT forecasting pipeline code from requirements. Use when scaffolding a pipeline with validation, logging, visualization, and repeatable runs. Trigger with "create TimeGPT pipeline", "build TimeGPT integration", or "generate forecast code".
agent-platform-model-registry
Agent Platform Model Registry Management. Use when you need to upload, list, describe, update, or delete machine learning models (and their versions) in the Agent Platform Model Registry. Don't use for model training, model deployment to endpoints, or managing non-Agent Platform models.