cost-metering-skill
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill manages local files (~/.claude/daily-cost.json and ~/.claude/cost-log.jsonl) to store and report on API usage metrics. These operations are limited to the user's home directory and do not involve sensitive system files or credentials.
- [INDIRECT_PROMPT_INJECTION]: The skill reads data from local JSON and JSONL files to generate reports. While this presents a potential ingestion surface for untrusted data if the log files were compromised by another process, the risk is minimal as the agent uses structured parsing (jq).
- Ingestion points: ~/.claude/daily-cost.json, ~/.claude/cost-log.jsonl
- Boundary markers: Not present.
- Capability inventory: Reads/writes local files, executes shell commands (cat, jq, mkdir, date).
- Sanitization: Data is processed as structured JSON via jq.
Audit Metadata