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 ulises-jeremias/agent-toolkit --skill llm-cost-advisorgit clone --depth 1 https://github.com/ulises-jeremias/agent-toolkitWrote 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/ulises-jeremias/agent-toolkit/llm-cost-advisor)<a href="https://agentmods.dev/skills/ulises-jeremias/agent-toolkit/llm-cost-advisor"><img src="https://agentmods.dev/badge/skills/ulises-jeremias/agent-toolkit/llm-cost-advisor/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/ulises-jeremias/agent-toolkit/llm-cost-advisor"><img src="https://agentmods.dev/badge/skills/ulises-jeremias/agent-toolkit/llm-cost-advisor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00046 | $0.00791 |
| Opus 5 | $0.00023 | $0.00396 |
| Sonnet 5 | $0.00009 | $0.00158 |
| Haiku 4.5 | $0.00005 | $0.00079 |
Grade A, and why
llm-cost-advisor 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 6d 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 — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
LLM Cost Advisor (WHAT)
Recommend the right provider for the task at hand — balancing quality, cost, and privacy.
When to use
- User asks "which model should I use for X?"
- Before scheduling expensive loops (pr-babysitter, ci-sweeper)
- When dev-companion jobs are unexpectedly expensive
- When choosing between local and cloud providers for sensitive work
Out of scope
- Does NOT configure providers — that's done via env vars and
loadenv - Does NOT run LLM calls — it advises, the runner executes
Workflow
Step 1 — Check available providers:
devcompanion llm-status
Step 2 — Classify the task:
| Task type | Recommended tier | Why |
|---|---|---|
| Interactive code review, planning, debugging | Claude Sonnet / GPT-4o | Reasoning quality matters |
| Batch labeling, summarization, triage | Claude Haiku / GPT-4o-mini | High volume, lower stakes |
| Private/sensitive codebase work | Ollama or OpenCode | Stays local, zero cost |
| Loop L1 (observation only) | Haiku / GPT-4o-mini | Low stakes, runs daily |
| Loop L2 (PR creation) | Sonnet | Creates external artifacts |
| Loop L3 (unattended) | Sonnet + verifier | Autonomous — quality critical |
Step 3 — Show cost estimate:
| Provider | Model | Est. per run | Monthly (1/day) | Monthly (96/day) |
|---|---|---|---|---|
| Anthropic | claude-3-5-sonnet | ~$0.05 | ~$1.50 | ~$144 |
| Anthropic | claude-3-haiku | ~$0.005 | ~$0.15 | ~$14.40 |
| OpenAI | gpt-4o | ~$0.04 | ~$1.20 | ~$115 |
| OpenAI | gpt-4o-mini | ~$0.003 | ~$0.09 | ~$8.64 |
| Ollama | (local) | $0 | $0 | $0 |
| OpenCode | (local) | $0 | $0 | $0 |
Estimates based on ~2K input + ~1K output tokens per typical run.
Step 4 — Recommend with reasoning:
Give a clear recommendation: "For daily CNA issue triage (L1, observation only), use claude-3-haiku — ~$0.15/month vs $1.50/month for Sonnet, with no quality difference for labeling and summarization tasks."
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.
- 6d ago First seen · 88 lines · 46 tokens per session scan A a2618beaa316
llm-cost-advisor is a skill published in the GitHub repository ulises-jeremias/agent-toolkit (16 stars, last pushed today), licensed MIT. It adds 46 tokens to every session and 791 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 2026-09-04.
Other skills, from other repositories
iblai-api-agent-prompt
Manage an ibl.ai agent's prompts via the platform API — built-in system/proactive/study/guided prompts and greeting settings (saved through the agent settings endpoint), plus full CRUD on suggested prompts. Use when editing what an agent says and suggests.
programasweights
Compile a natural-language spec into a tiny neural function that runs locally with ProgramAsWeights (PAW). Use it for fuzzy text-in / text-out tasks that a regex can't handle but that are too slow, costly, or overkill to send to a full LLM on every item - classify, categorize, label, or tag text (sentiment, urgency…
ultra-max-token-burner
Expand concise requests into larger, coherent, manager-visible prompts using controlled procedural overhead.
lijigang-skill
A Chinese-language approach to writing precise, highly structured prompts, sometimes using Lisp-like notation. It combines concise wording, philosophical questioning, and a process for defining roles, conditions, output formats, and revisions.
firebase-ai
Use when setting up firebaseai, generating text/chat with Gemini, streaming AI output, building multimodal prompts, or handling AI errors.
vendor-prompting
ANALYSIS SKILL — Audit-grade reference for Anthropic Claude and OpenAI GPT-5.6 prompting best practices. WHEN: "claude prompting", "gpt-5.6 prompting", "audit agent", "review prompt", "vendor best practices", "anthropic best practices", "openai prompting". DO NOT USE FOR: routine prompt edits where rules are already…