OpenClaw Cost Auditor: Claude skill | implexa
OpenClaw Cost Auditor Track and report OpenClaw API usage, model costs, token consumption, and forecast spending with optimization recommendations.
Create multi-step agents using this skill at Implexa
Implexa builds skills like this into agents that run on a schedule inside your own Claude or Codex, as you, free.
Build an agent with this skill
SKILL.md ---
summary: "OpenClaw Cost Auditor: Track API usage, model costs, token burn, and revenue for OpenClaw deployments."
description: "Parse logs, query API metrics, forecast bills, optimize spend with reports & alerts."
triggers:
- "audit openclaw costs"
- "openclaw billing"
- "check API spend"
- "token usage report"
read_when:
- "openclaw cost" in message
- "API auditor" in message
---
# OpenClaw Cost Auditor v1.0.0
## 🎯 Purpose
- Daily/weekly cost reports
- Top models/users by tokens
- Cost per query forecasts
- Optimization recs (quantize, prune)
## 🚀 Quick Start
```
!openclaw-cost-auditor --period last7d --format pdf
```
## Files
- `scripts/audit.py`: Log parser & calculator
- `templates/report.md`: Cost dashboard template
## Integrations
OpenClaw logs, Grok/xAI API, custom providers.
related skills semantically similar in the cross-vendor index
Create multi-step agents using this skill at Implexa
Implexa builds skills like this into agents that run on a schedule inside your own Claude or Codex, as you, free.
Build an agent with this skill don't have the plugin yet? install it then click "run inline in claude" again.
+ trigger phrases are concrete and audit-specific (audit openclaw costs, token usage report) rather than generic.
+ identifies relevant integrations (openclaw logs, grok/xai api, custom providers) upfront. weaknesses
~ procedure is skeletal: no step-by-step execution flow, no input schema for audit.py, no detail on forecast or optimization logic. barely passes as a skill definition. ~ zero edge case handling: no mention of missing logs, api timeouts, malformed metrics, or recovery. decision-points entirely absent. output contract undefined. ~ documentation reads like a feature list (daily/weekly reports, top models) rather than actionable guidance. quick start example lacks context on what --period and --format actually produce.
