CostLens
v1.0.1Calculate OpenClaw usage cost from offline event logs, apply budget thresholds, and export operator-facing reports.
MIT-0
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The skill declares a Node.js CLI that reads offline JSON event files, applies local pricing, checks budgets, and exports reports — and the code, README, SKILL.md, fixtures, and tests all implement exactly that. Required binaries (node) are appropriate and proportional.
Instruction Scope
SKILL.md instructs the agent to run the local Node CLI against a local events JSON file and optionally write a report file. The runtime instructions and the implemented code only read local files, compute aggregates, and write local outputs; they do not request additional system files, environment variables, or remote endpoints.
Install Mechanism
No install spec is provided (instruction-only + included source files). This minimizes install-time risk. The repository contains only local Node source, no downloads or external installers.
Credentials
The skill declares no required environment variables, credentials, or config paths, and the code does not read environment secrets. All data inputs are explicit file paths (events, out) supplied by the user.
Persistence & Privilege
The skill does not request always:true and does not modify other skills or global agent settings. It only writes report files to paths supplied by the user.
Assessment
This skill appears to do what it says: local offline analysis of JSON token events using Node.js. Before installing/running, ensure you have Node 18+ and run it only on event files you trust because it will read the events JSON and write report files to the paths you provide. It does not contact external services or require credentials. If you plan to let an autonomous agent invoke this skill, remember it can read any events file path you pass to it — avoid passing files that contain secrets or unrelated data.Like a lobster shell, security has layers — review code before you run it.
budgetcostlatestllmmonitoringopenclaw
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
💸 Clawdis
Binsnode
SKILL.md
CostLens
Turn JSON usage events into cost summaries, budget checks, and offline reports.
When to use
- You already have token usage events and need a quick spend summary.
- You want a simple budget gate in local automation or CI.
- You need an exportable JSON report without relying on a hosted billing dashboard.
Commands
node {baseDir}/bin/costlens.js monitor --events ./events.json --budget 10.00 --threshold 80
node {baseDir}/bin/costlens.js budget check --events ./events.json --budget 5.00 --format json
node {baseDir}/bin/costlens.js report --events ./events.json --out ./reports/cost-report.json
Built-in default rates
| Model | Input/1k | Output/1k |
|---|---|---|
| gpt-4.1 | $0.01 | $0.03 |
| gpt-4o-mini | $0.00015 | $0.0006 |
| claude-3-5-sonnet | $0.003 | $0.015 |
| default | $0.002 | $0.008 |
Events can override the defaults with inputCostPer1k and outputCostPer1k.
Event shape
[
{
"model": "gpt-4.1",
"promptTokens": 1500,
"completionTokens": 800,
"timestamp": "2026-02-26T10:00:00Z"
}
]
Output
- Total calls, total tokens, and total cost
- Per-model breakdowns for calls, tokens, and cost
- Per-day spend trends
- Budget usage percentage and alert level (
ok,warning, orcritical)
Boundaries
- Pricing is based on built-in defaults plus per-event overrides. It is not a live pricing feed.
- CostLens is optimized for offline JSON analysis, not streaming metering or invoice reconciliation.
Files
9 totalSelect a file
Select a file to preview.
Comments
Loading comments…
