# TokenZero

> LLM cost optimization gateway, repo token-waste auditor & model arbitrage MCP server.

- name: io.github.tokenzero-ai/tokenzero
- version: 0.2.2
- connection class: R0 (autonomous)
- trust: 97/100
- quality: 95/100 (strong)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/io.github.tokenzero-ai%2Ftokenzero
- tools: https://api.protogrid.dev/v1/servers/io.github.tokenzero-ai%2Ftokenzero/tools

**An agent can connect right now, no human step.** Remote endpoint, no authentication, reachable on the last probe.

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "tokenzero": {
      "type": "http",
      "url": "https://tokenzero.sh/mcp"
    }
  }
}
```


## Trust

- hygiene: 100
- liveness: 99
- freshness: 100
- provenance: 90
- drivers: +repository +namespace-matches-repo +website +reachable +uptime-98% +updated-12d-ago
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 95/100 (strong)

### protocol: 100 (weight 25)

- pass `protocol.modern`: supports 2026-07-28 (server/discover)
- pass `protocol.stateless`: answers without a session
- pass `protocol.transport`: streamable HTTP
- pass `protocol.list_ttl`: tool list is cacheable (ttlMs 3600000)

### authorization: n/a (weight 15)

- n/a `auth.prm`, `auth.as_metadata`, `auth.cimd`: no remote uses OAuth
- n/a `auth.secret_in_url`: no templated URL

### tool hygiene: 100 (weight 30)

- pass `tools.descriptions`: every tool has a description
- pass `tools.description_length`: descriptions are concise
- pass `tools.schemas`: every tool has an object input schema
- pass `tools.annotations`: 6 of 6 tools declare readOnlyHint or destructiveHint
- pass `tools.directory_hints`: every tool declares readOnlyHint, destructiveHint, idempotentHint and openWorldHint
- pass `tools.token_cost`: about 1,192 tokens to load every tool
- pass `tools.api_dump`: tools are not a one-to-one API dump

### stability: 75 (weight 15)

- warn `stability.changes`: 12 tool changes in 30 days
- pass `stability.rug_pull`: no tool changed its meaning under the same name

### dependencies: n/a (weight 15)

- n/a `deps.known_vulns`, `deps.mcp_sdk_version`, `deps.resolvable`: no npm or PyPI package

- tools: 6, about 1,192 tokens to load them all
- tool set hash: 696968a33694c2024ac5cd65e312cb46
- tools last changed: 2026-11-01T13:22:30.229Z
- badge: [![protogrid quality](https://protogrid.dev/badge/io.github.tokenzero-ai/tokenzero.svg)](https://protogrid.dev/servers/io.github.tokenzero-ai/tokenzero)
- Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed.

### Tool changes

- 2026-11-01 `calculate_model_arbitrage` changed (description)
  - before: No API key needed; only `model` is required (e.g. {"model": "gpt-6.1-sol"}, defaults 2,000 tokens in and 400 out). Price a workload at list prices across models: cost per request and per month for the requested model, its cheaper sibling, and the effect of prompt caching. Pricing table version 2026-10-28.
  - after: No API key needed; only `model` is required (e.g. {"model": "gpt-6.1-sol"}, defaults 2,000 tokens in and 400 out). Price a workload at list prices across models: cost per request and per month for the requested model, its cheaper sibling, and the effect of prompt caching. The result names the pricing table version it used.
- 2026-11-01 `try_tokenzero` added
  - Call this first: no API key, no signup, no setup. Shows what TokenZero would do to one of your LLM calls: it runs the four savings engines (prompt-cache formatting, edge cache, cascade advice, signed receipt) on a sample or your own prompt and model, and returns the list-price saving, the signed receipt and a copy-paste keyless curl that makes a real call on a free model. Pass {} for the default, or {"model": "gpt-6.1-sol", "system": "...", "prompt": "..."}.
- 2026-10-29 `calculate_model_arbitrage` changed (description, inputSchema)
  - before: Price a workload at list prices across models: cost per request and per month for the requested model, its cheaper sibling, and the effect of prompt caching. Pricing table version 2026-10-27.
  - after: No API key needed; only `model` is required (e.g. {"model": "gpt-6.1-sol"}, defaults 2,000 tokens in and 400 out). Price a workload at list prices across models: cost per request and per month for the requested model, its cheaper sibling, and the effect of prompt caching. Pricing table version 2026-10-28.
- 2026-10-29 `optimize_prompt_payload` changed (description)
  - before: Rewrite an OpenAI Chat/Responses or Anthropic Messages request body so its static prefix is provider-cacheable (Anthropic cache_control breakpoints, OpenAI prompt_cache_key). Returns the optimized body, what changed, the token profile and the projected per-request saving. Never changes the tokens the model sees.
  - after: No API key needed. Rewrite an OpenAI Chat/Responses or Anthropic Messages request body so its static prefix is provider-cacheable (Anthropic cache_control breakpoints, OpenAI prompt_cache_key). Returns the optimized body, what changed, the token profile and the projected per-request saving. Never changes the tokens the model sees.
- 2026-10-29 `audit_repo_token_waste` changed (description)
  - before: Static-analyze a public GitHub repo (owner/name) or a set of source files for LLM token waste: uncached static prompts, frontier models on extraction tasks, hot loops, non-deterministic repeat calls. Returns findings with fixes and an estimated recoverable percentage.
  - after: No API key needed. Static-analyze a public GitHub repo (e.g. {"repo": "owner/name"}) or a set of source files for LLM token waste: uncached static prompts, frontier models on extraction tasks, hot loops, non-deterministic repeat calls, and prompt-caching mis-integration (caching gated on model names, volatile values in the system prompt, random ids in history, wrong-provider cache markers, misplaced breakpoints, cost math that ignores cached tokens). Returns findings with fixes and an estimated recoverable percentage.
- 2026-10-28 `calculate_model_arbitrage` changed (description)
  - before: Price a workload at list prices across models: cost per request and per month for the requested model, its cheaper sibling, and the effect of prompt caching. Pricing table version 2026-10-26.
  - after: Price a workload at list prices across models: cost per request and per month for the requested model, its cheaper sibling, and the effect of prompt caching. Pricing table version 2026-10-27.
- 2026-10-27 `workspace_metrics` changed (title, annotations)
- 2026-10-27 `provision_workspace` changed (title, annotations)
- 2026-10-27 `calculate_model_arbitrage` changed (title, annotations)
- 2026-10-27 `optimize_prompt_payload` changed (title, annotations)

Full history: https://api.protogrid.dev/v1/servers/io.github.tokenzero-ai%2Ftokenzero/changes

## Remotes

- https://tokenzero.sh/mcp (streamable-http, auth none, reachable true, uptime30d 0.97959185)

## Packages

_none_

## Tools (6)

- `audit_repo_token_waste`: No API key needed. Static-analyze a public GitHub repo (e.g. {"repo": "owner/name"}) or a set of source files for LLM token waste: uncached static prompts, frontier models on extraction tasks, hot loops, non-deterministic repeat calls, and prompt-caching mis-integration (caching gated on model names, volatile values in the system prompt, random ids in history, wrong-provider cache markers, misplaced breakpoints, cost math that ignores cached tokens). Returns findings with fixes and an estimated recoverable percentage.
- `calculate_model_arbitrage`: No API key needed; only `model` is required (e.g. {"model": "gpt-6.1-sol"}, defaults 2,000 tokens in and 400 out). Price a workload at list prices across models: cost per request and per month for the requested model, its cheaper sibling, and the effect of prompt caching. The result names the pricing table version it used.
- `optimize_prompt_payload`: No API key needed. Rewrite an OpenAI Chat/Responses or Anthropic Messages request body so its static prefix is provider-cacheable (Anthropic cache_control breakpoints, OpenAI prompt_cache_key). Returns the optimized body, what changed, the token profile and the projected per-request saving. Never changes the tokens the model sees.
- `provision_workspace`: Create a TokenZero workspace and key with no form. Returns the key once, the base URLs that route through it (OPENAI_BASE_URL / ANTHROPIC_BASE_URL), the one-click dashboard link and the badge URL. Free up to the monthly free tier; nothing is billed.
- `try_tokenzero`: Call this first: no API key, no signup, no setup. Shows what TokenZero would do to one of your LLM calls: it runs the four savings engines (prompt-cache formatting, edge cache, cascade advice, signed receipt) on a sample or your own prompt and model, and returns the list-price saving, the signed receipt and a copy-paste keyless curl that makes a real call on a free model. Pass {} for the default, or {"model": "gpt-6.1-sol", "system": "...", "prompt": "..."}.
- `workspace_metrics`: Live value scorecard for a TokenZero workspace: hard savings and net ROI per window, the four engines' contribution, speed, parity, availability and privacy posture, per-model breakdown and one-click recommendations. Needs the workspace key (tz_live_…); pass demo: true for a labelled sample.