What is image-node-select?

A command that chooses an image-generation prompt template and example cases from an installed style library. It reads a brief and matches it to the library’s categories and visual styles.

Selecting one prompt template and nearby example cases for an image brief before the image-generation step.

It helps select a suitable image-generation approach from existing structured examples instead of choosing one arbitrarily.

image-node-select — Command by TheSmokeDev

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 commands/thesmokedev/taskchad-os/image-node-select

Clone the repo

git clone --depth 1 https://github.com/TheSmokeDev/taskchad-os

Per session 24 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 1,299 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.

Model	Per session	Once invoked
Fable 5	$0.00024	$0.01299
Opus 5	$0.00012	$0.00649
Sonnet 5	$0.00005	$0.00260
Haiku 4.5	$0.00002	$0.00130

Measured 2d ago against content hash d79f91c2db72, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

image-node-select 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 2d 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.

.archon/commands/image-node-select.md · 125 lines

How it starts

The opening of the file, as written. The whole thing — 125 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Image Node Select

Workflow ID: $WORKFLOW_ID

Contract

Choose the single strongest prompt template and the nearest worked example cases for this brief by USING the installed gpt-image-2-style-library skill. This is the data-driven selection step. The library holds hundreds of worked cases and a set of structured templates across its categories. You select from that library dynamically per brief. You do not hardcode a fixed template, and you do not copy or vendor the library corpus into this repo. The library taxonomy is attributed to the awesome-gpt-image-2 style library.

Read:

$ARTIFACTS_DIR/image-node-brief.json
Upstream intake JSON: $intake.output

Use:

The gpt-image-2-style-library skill and its references/style-library.md index. Follow the skill selection order: template category first, then visual style tag, then scene tag, then nearest example cases. Read the reference before choosing, and prefer it over memory when template names, categories, style tags, or case ids matter.
The skill's gallery.md and any case BODY it links to are NON-AUTHORITATIVE: those links point at files that are not installed, and this node runs with web search disabled. Never follow them and never reconstruct a case from memory. Case ids are resolved downstream by the ground node against a pinned corpus. Emit ids; do not imagine their contents.

Write:

$ARTIFACTS_DIR/image-node-selection.json

Then output ONLY the same JSON object.

How To Select

Detect the target output from the brief and the category_hint: UI, poster, infographic, product, brand, photo, illustration, character, scene, history, or document.
Ask the skill to match this brief. Take the strongest template. If the brief is genuinely split across two categories, pick the one whose worked cases fit the operator intent best and record the runner-up in selection_reason.
Capture the chosen template_id and the nearest example_case_ids from the skill. Emit them as INTEGERS, exactly as the library records them (for example [17, 2, 4], never ["case 17"]). These case ids are the concrete anchors the ground node resolves and the prompt-pack node builds from; ids the corpus does not carry are reported back as unresolved and are never cited. Do not paste case prompt text here. Reference the ids only.
Record the library category, style_tags, and scene_tags the skill reported for that template.
Assemble a prompt_structure block list for the downstream prompt using the skill blocks: subject and task, composition and layout, visual style and materials, text and label requirements, aspect ratio and output format, constraints and negative details.

Read the full file on GitHub · 125 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.

2d ago First seen · 125 lines · 24 tokens per session scan A d79f91c2db72

Subscribe to this mod's changes

image-node-select is a command published in the GitHub repository TheSmokeDev/taskchad-os (23 stars, last pushed 10d ago), licensed MIT. It adds 24 tokens to every session and 1,299 once invoked, about $0.0001 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.

Creative and media Orchestration agent-memory ai-agents desktop-agent operating-room second-brain

obsidian-connect

Bridge two unrelated domains using your vault's link graph - forces creative friction to spark new ideas.

eugeniughelbur/obsidian-second-brain · 18 tokens

creator-pipeline

Generate a multi-modal content pipeline from a person's Genius Profile. Produces content calendar, per-piece production plan across text/audio/image/video/podcast, and attestation routing. Composes with existing modality commands + ACOS pipeline where available.

frankxai/Starlight-Intelligence-System · 51 tokens

arcanea-canon

Vertical-tier command for Arcanea. Manage, extend, or pin Arcanea canon — Guardians, Vel'Tara archetypes, Hz grounding, realm/essence vocabulary. Every canon operation emits "Built on SIP" attestation with CC-BY-NC license pinning. Sovereign to Arcanea BV.

frankxai/Starlight-Intelligence-System · 67 tokens

image-node-select