coreclaw

Agent readability score

Strengths

• Most tool names clearly indicate their purpose (e.g., 'list_proxy_regions', 'run_worker')
• Descriptions include specific 'WHEN TO USE' scenarios for context-aware selection

Weaknesses

• Chinese trigger text is identical across 20+ tools, creating redundancy
• Multiple tools with similar names (e.g., 'rerun_last_worker_run' vs 'rerun_worker_last_run')
• Lack of explicit error handling documentation in tool descriptions

• list_proxy_regions — Chinese trigger text is identical to 20+ other tools
• rerun_last_worker_run — Functionally overlaps with 'rerun_worker_last_run'
• export_last_worker_run_results — Redundant with 'export_worker_run_results'
• get_worker — Vague name that could refer to multiple worker-related actions

Verdict

Tools provide clear intent through naming and context-specific 'WHEN TO USE' sections, but significant overlap between similar functions and repetitive Chinese triggers reduce usability. The lack of explicit error handling documentation further lowers the readability score despite functional clarity.

9/11/2026, 2:57:48 AM

Event narrative

coreclaw moved from 'auth-gated' to 'legacy' era, achieving 100 compliance score.

Risk: medium

• 2026-09-10T21:16:35.016Z — era changed from 'auth-gated' to 'legacy' · transition to legacy era may indicate stability or policy shift
• 2026-09-11T01:59:45.474Z — complianceScore improved from null to 100 · perfect compliance achieved, suggesting system health or configuration change

Advice

Monitor complianceScore trends for consistency. Verify if 'legacy' era alignment affects future operations. Ensure smokeOk is addressed when available.