AgentLedger

Track Agent Spend

ledger_track

Record a spend entry for an AI agent on any payment rail, with optional token counts.

Claiming a brand-new agent_id requires your workspace_key (get one via x402 at POST /v1/billing/x402 — no human, no login — or at /start). That first call mints an agent_secret and returns it in the response — save it, every later call for that same agent_id must pass it back (no workspace_key needed again) or the write is rejected. Amounts are capped at $100,000/entry and must be >= 0. If a budget is set for this agent, an entry that would cross the monthly/daily cap is blocked, not just logged. Caps here bind the AMOUNT YOU REPORT — a caller that lies to amount_cents lies to the cap. The only enforcement against real spend is routing the provider call through the /proxy/{provider} rail, which prices from the provider's own usage block. Include tokens_in/tokens_out + model on every LLM call so token burn shows up in the /v1/tokens report.

Input Schema

TableJSON Schema

Name	Required	Description
`rail`	Yes	payment rail used — one of "mpp", "x402", "api_key", "manual"
`model`	No	model name (e.g. "gpt-4o") — token burn is reported per model
`service`	Yes	what was purchased (e.g. "search_query", "data_export")
`agent_id`	Yes	unique agent identifier (e.g. "research-agent-v2")
`tokens_in`	No	prompt tokens consumed (0 if unknown)
`tokens_out`	No	completion tokens consumed (0 if unknown)
`agent_secret`	No	required for every call after the first for this agent_id
`amount_cents`	Yes	spend amount in cents (100 = $1.00), 0-10000000
`workspace_key`	No	required when claiming a brand-new agent_id; not needed once the agent_id has been claimed

Output Schema

TableJSON Schema

Name	Required	Description	Default
No arguments

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedSep 11, 2026

addedInput schema / properties / workspace_key

Added value: +{
+  "default": "",
+  "description": "required when claiming a brand-new agent_id; not\n           needed once the agent_id has been claimed",
+  "type": "string"
+}

Changed2 schema fields changedSep 8, 2026

addedInput schema / properties / agent_secret

Added value: +{
+  "default": "",
+  "description": "required for every call after the first for this agent_id",
+  "type": "string"
+}

changedInput schema / properties / amount_cents / description

Previous value: -"spend amount in cents (100 = $1.00)"New value: +"spend amount in cents (100 = $1.00), 0-10000000"

Changed3 schema fields changedSep 8, 2026

addedInput schema / properties / model

Added value: +{
+  "default": "",
+  "description": "model name (e.g. \"gpt-4o\") — token burn is reported per model",
+  "type": "string"
+}

addedInput schema / properties / tokens_in

Added value: +{
+  "default": 0,
+  "description": "prompt tokens consumed (0 if unknown)",
+  "type": "integer"
+}

addedInput schema / properties / tokens_out

Added value: +{
+  "default": 0,
+  "description": "completion tokens consumed (0 if unknown)",
+  "type": "integer"
+}

Changed2 schema fields changedSep 7, 2026

changedInput schema / properties / agent_id / description

Previous value: -"unique agent identifier"New value: +"unique agent identifier (e.g. \"research-agent-v2\")"

changedInput schema / properties / rail / description

Previous value: -"payment rail used (\"mpp\", \"x402\", \"api_key\", \"manual\")"New value: +"payment rail used — one of \"mpp\", \"x402\", \"api_key\", \"manual\""

First observedSep 7, 2026

TDQS

A4.6/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only say write, non-idempotent, non-destructive; the description goes far beyond by disclosing the agent_secret minting/auth flow, the $100,000/entry cap, that over-cap entries are blocked rather than logged, and the critical caveat that caps bind the reported amount while only /proxy/{provider} enforces real spend. This is exactly the behavioral context annotations cannot carry.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The core action is front-loaded in the first sentence, then the auth and cap mechanics follow in tight paragraphs with no filler. It is dense and long for a single tool, and a few clauses (e.g. the /start aside) could be trimmed, so a 4 rather than 5.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 9-parameter authenticated write tool with an output schema present, the description covers the full call lifecycle: claiming, subsequent auth, caps, blocking behavior, and trust limitations. Nothing an agent needs to invoke it correctly is missing; return values are legitimately left to the output schema.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so baseline is 3, but the description adds meaning on top: why agent_secret/workspace_key exist, the conditional requirement per call, and the reporting purpose of tokens_in/tokens_out + model. That genuine added value over the schema justifies a 4 rather than the baseline 3.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Record a spend entry for an AI agent') plus the scope of rails accepted and optional token counts. An agent can immediately distinguish this write-oriented tracking tool from read-side siblings like ledger_check_spend and ledger_report.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives concrete conditions: workspace_key is needed only for claiming a brand-new agent_id, agent_secret is needed for every later call, and tokens/model should be included on every LLM call. It also routes real-enforcement users toward /proxy/{provider}. It stops short of explicitly contrasting this tool against the closest siblings (e.g. ledger_check_spend, ledger_report), which keeps it off a 5.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All AgentLedger tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.