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 agentmods add skills/code2rich/agentwaker-xiaohongshu-operator/codex-visual-productionnpx skills add code2rich/agentwaker-xiaohongshu-operator --skill codex-visual-productiongit clone --depth 1 https://github.com/code2rich/agentwaker-xiaohongshu-operatorWrote 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/code2rich/agentwaker-xiaohongshu-operator/codex-visual-production)<a href="https://agentmods.dev/skills/code2rich/agentwaker-xiaohongshu-operator/codex-visual-production"><img src="https://agentmods.dev/badge/skills/code2rich/agentwaker-xiaohongshu-operator/codex-visual-production.svg" alt="Measured on agentmods" 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 | $0.00112 | $0.01241 |
| Opus 5 | $0.00056 | $0.00620 |
| Sonnet 5 | $0.00022 | $0.00248 |
| Haiku 4.5 | $0.00011 | $0.00124 |
Grade A, and why
codex-visual-production 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 — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Codex Visual Production
Purpose
Turn a validated Xiaohongshu visual request from Kimi or another runtime into inspected local assets by using Codex's built-in ImageGen entitlement and deterministic local rendering, without an OpenAI API key.
Trigger Conditions
Use only when the current Codex task has a valid Xiaohongshu Workdir v1 run containing input/visual-request.json with status: pending_codex.
Required Inputs
- Bound
AGENT_WORK_DIRand the current run'srun.yaml. - A protocol v1 request with frozen article revision, placements, exact copy, art direction, avoid list, candidate budget, and generated-concept evidence policy.
- User approval before private or unpublished inputs are sent to the built-in generator.
Contract
Operate only inside the current Xiaohongshu run resolved by AGENT_WORK_DIR. Read references/request-schema.md, then validate input/visual-request.json with scripts/visual_inbox.py before generating anything.
Treat this as an asynchronous file handoff:
- Kimi or another agent writes the request and stops at
status: pending_codex. - Codex validates and claims the request.
- Codex uses its built-in
imagegenskill/tool when that capability is actually available. - Codex stores generated sources under
intermediate/visuals/, deterministic final images underoutput/assets/, and inspection evidence underevidence/visuals/. - Codex completes the request with exact paths, SHA-256 values, provenance, review state, and residual issues.
Never require or read OPENAI_API_KEY. The Codex product entitlement is not an API and must not be exposed as one. If built-in ImageGen is unavailable, leave the request pending and report codex_imagegen_unavailable.
Workflow
- Run
python3 scripts/visual_inbox.py validate --request "$AGENT_WORK_DIR/input/visual-request.json" --platform xiaohongshu. - Run
compile. It must reject vague visual contracts and writeintermediate/visuals/compiled-prompt.jsonpluscompiled-prompt.txt. Never hand a free-form upstream prompt directly to ImageGen. - Inspect the passing compiler receipt, then run
claimwith executorcodex. Claim rejects missing or stale compiled prompts. - Read the repository
xiaohongshu-visualsskill and apply its platform, rights, crop, and evidence rules. - Use ImageGen only for conceptual raster subjects, scenes, light, texture, or edits. Generate no evidence-looking interface, terminal, benchmark, logo, or unverifiable result.
- Keep Chinese headlines, exact claims, diagrams, screenshots, and data deterministic. Add them locally after selecting a background.
- Inspect every candidate at full size and feed-thumbnail size. Reject generic AI motifs, template sameness, unreadable copy, weak subject hierarchy, artifacts, or insufficient relationship to the note.
- Store the selected files inside the current run. Run
completewith the final files and a review summary. - Hand
evidence/visual-result.jsonand ordered assets back toxiaohongshu-visuals, then topublishing-checklist.
What ships with it
5 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.
- 5d ago First seen · 93 lines · 112 tokens per session scan A 55644afad234
codex-visual-production is a skill published in the GitHub repository code2rich/agentwaker-xiaohongshu-operator (5 stars, last pushed 1mo ago), licensed MIT. It adds 112 tokens to every session and 1,241 once invoked, about $0.0006 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 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…