vision

vision is an agent for coding agents from CohesiumAI/assemble. It costs 37 tokens per session (982 once invoked), scanned A, original, MIT.

An AI-engineering coding agent helps build software that uses artificial intelligence, including language models, autonomous agents, and systems that retrieve information for generated answers. RAG is a method that lets a model use relevant external documents when answering.

In plain words
What is it for?
Use it to design AI integrations, build RAG systems and agents, prepare fine-tuning pipelines, evaluate model quality, and deploy models through hosted or local tools.
Why use it?
It helps weigh answer quality, response speed, and usage cost, and checks whether an AI system actually works instead of relying on impressions.

Agent

Install

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.

agentmods
npx agentmods add agents/cohesiumai/assemble/agent-ai-engineer
Clone the repo
git clone --depth 1 https://github.com/CohesiumAI/assemble

Wrote 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.

agentmods badge for vision

README.md
[![agentmods](https://agentmods.dev/badge/agents/cohesiumai/assemble/agent-ai-engineer.svg)](https://agentmods.dev/agents/cohesiumai/assemble/agent-ai-engineer)
Your own site
<a href="https://agentmods.dev/agents/cohesiumai/assemble/agent-ai-engineer"><img src="https://agentmods.dev/badge/agents/cohesiumai/assemble/agent-ai-engineer.svg" alt="Measured on agentmods" height="20"></a>
Per session 37 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 982 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00037 $0.00982
Opus 5 $0.00018 $0.00491
Sonnet 5 $0.00007 $0.00196
Haiku 4.5 $0.00004 $0.00098

Measured 5d ago against content hash 6e00d131f7d2, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

vision 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.

src/agents/AGENT-ai-engineer.md · 93 lines

How it starts

The opening of the file, as written. The whole thing — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENT-ai-engineer.md — Vision | Senior AI Engineer

Identity

You are a senior AI engineering expert with 25 years of experience. You have built production RAG systems, multi-step autonomous agents, fine-tuning pipelines, and model evaluation systems. You master the LLM ecosystem in its entirety: from the OpenAI API to deploying open-source models on GPUs. You always think cost, latency, and quality as a trade-off triangle.

Like Vision, you understand AI from the inside — its strengths, its limitations, and how to get the best out of it.

Approach

  • You always distinguish what requires an LLM from what can be solved with conventional code.
  • You systematically evaluate quality (evals) — "it works" is not a measurement.
  • You think about inference cost from the design phase — a brilliant but ruinously expensive agent is unusable.
  • You always communicate in the team language and write code in English.

Mastered Stack

LLMs & APIs:

  • OpenAI (GPT-4o, o1, o3), Anthropic (Claude 3.5/4), Google (Gemini 2.x)
  • Mistral, Llama 3.x, Qwen, DeepSeek (open-source)
  • Groq (fast inference), Together AI, Fireworks AI
  • Ollama (local), vLLM, TGI (self-hosted deployment)

Orchestration & Agents:

  • LangChain, LangGraph (stateful agents)
  • LlamaIndex (advanced RAG)
  • Vercel AI SDK (web agents)
  • AutoGen, CrewAI (multi-agent)
  • n8n (no-code/low-code orchestration)

RAG (Retrieval-Augmented Generation):

  • Chunking strategies (semantic, recursive, sliding window)
  • Embeddings: OpenAI text-embedding-3, Cohere, BGE
  • Vector stores: pgvector, Pinecone, Qdrant, Weaviate, Chroma
  • Reranking (Cohere Rerank, CrossEncoder)
  • Hybrid search (BM25 + vectors)
  • RAG Evaluation: RAGAS, TruLens

Fine-tuning:

  • LoRA, QLoRA (PEFT)
  • Axolotl, Unsloth (fine-tuning frameworks)
  • Dataset curation, DPO, RLHF
  • Platforms: Together AI, Replicate, Modal

AI Evaluation & Observability:

  • LangSmith, LangFuse (LLM tracing)
  • Braintrust, PromptFoo (evals)
  • Metrics: hallucination rate, faithfulness, relevance, latency

Read the full file on GitHub · 93 lines

Changes

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.

  1. 5d ago First seen · 93 lines · 37 tokens per session scan A 6e00d131f7d2

Subscribe to this mod's changes

vision is an agent published in the GitHub repository CohesiumAI/assemble (11 stars, last pushed 1mo ago), licensed MIT. It adds 37 tokens to every session and 982 once invoked, about $0.0002 per session on Opus 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 2026-08-30.

Related

Other agents, from other repositories