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 tushaarmehtaa/tushar-skills --skill ai-cost-auditgit clone --depth 1 https://github.com/tushaarmehtaa/tushar-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/tushaarmehtaa/tushar-skills/ai-cost-audit)<a href="https://agentmods.dev/skills/tushaarmehtaa/tushar-skills/ai-cost-audit"><img src="https://agentmods.dev/badge/skills/tushaarmehtaa/tushar-skills/ai-cost-audit/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/tushaarmehtaa/tushar-skills/ai-cost-audit"><img src="https://agentmods.dev/badge/skills/tushaarmehtaa/tushar-skills/ai-cost-audit.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 | $0.00043 | $0.00919 |
| Opus 5 | $0.00022 | $0.00460 |
| Sonnet 5 | $0.00009 | $0.00184 |
| Haiku 4.5 | $0.00004 | $0.00092 |
Grade A, and why
ai-cost-audit 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 12d 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 — 74 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI cost audit
Build a cost model from observed usage and current primary-source prices. Treat repository intent, measured usage, invoices, and projections as different evidence classes.
Choose a mode
- Inventory: map model/media calls, routing, retries, and ownership.
- Economics: calculate cost per action, user, plan, and month.
- Simulation: compare pricing, volume, model, cache, batch, or abuse scenarios.
- Optimization: rank changes after measuring quality and operational risk.
- Reconciliation: explain the gap between bottom-up estimates and provider invoices.
- Verification: confirm that a completed change reduced spend without unacceptable quality or latency regressions.
Use the narrowest mode that answers the request. Combine modes only when the user asks for a full audit or the dependency is necessary.
Workflow
- Inspect the repository, existing telemetry, billing exports, pricing configuration, and prior analyses before asking questions.
- State the audit boundary: environments, date range, providers, features, currencies, taxes, credits, and whether non-model infrastructure is included.
- Inventory direct and indirect calls: generation, reasoning, embeddings, reranking, tools, image/audio/video, moderation, retries, fallbacks, agents, queues, evaluations, and batches.
- Prefer provider-metered tokens or media units. Keep measured values, code-derived estimates, generic estimates, and assumptions visibly separate. Use p50, p95, and worst-case where available.
- Fetch current prices only from official provider sources when pricing affects the answer. Record URL, retrieval date, region/tier/currency, and special terms such as cached input, reasoning tokens, batch, storage, or minimum charges. Do not rely on bundled price tables or memory.
- Model each cost path, including failed calls, retry amplification, tool loops, cache writes/reads, storage, egress, gateway fees, payment fees, free allowances, and shared fixed costs when relevant.
- Reconcile the modeled total against invoices or billing dashboards. Quantify unexplained variance instead of forcing agreement.
- Run normal, high-usage, abuse, and sensitivity scenarios. Do not apply universal margin or traffic thresholds without the product's business constraints.
- Rank recommendations by expected savings range, evidence confidence, quality risk, latency effect, engineering effort, reversibility, and measurement plan.
- Require an evaluation and canary before changing models, prompts, routing, or output limits. Verify spend, quality, latency, error rate, and user outcomes afterward.
What ships with it
2 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.
- 12d ago First seen · 74 lines · 43 tokens per session scan A da762a1b7e28
ai-cost-audit is a skill published in the GitHub repository tushaarmehtaa/tushar-skills (11 stars, last pushed 2d ago), licensed MIT. It adds 43 tokens to every session and 919 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.
Other skills, from other repositories
sector-rotation
An analysis framework for comparing industries in the Chinese A-share stock market, using business conditions, price momentum, valuation, and money flows. It produces rankings and higher- or lower-allocation suggestions.
strategy-pivot-designer
Detect backtest iteration stagnation and generate structurally different strategy pivot proposals when parameter tuning reaches a local optimum.
twitter-reader
Read Twitter/X for financial research using opencli (read-only). Use this skill whenever the user wants to read their Twitter feed, search for financial tweets, view bookmarks, look up user profiles, or gather market sentiment from Twitter/X. Triggers include: "check my feed", "search Twitter for", "show my…
chenhao-limit-up
A framework for judging Chinese A-share stocks that have reached the daily price-rise limit, using market mood, sector leadership, and trading momentum.
furusato
A Japanese hometown-tax donation manager for furusato nozei, a system where donations to municipalities can qualify for an income-tax or local-tax deduction. It reads donation receipts, stores donation records, and calculates deduction limits.
reading-receipt
An image-reading workflow for extracting structured information from receipts, invoices, and hometown-tax donation certificates. It can first extract text from PDFs and otherwise read their images.