Claude Code Agent Harness | HarnessRouter
Browse 41 harnesses
30-day usage· Aug 12–Sep 10, 2026
Data as of September 11, 2026 · About the data
Rank #2#2 of21CLI agents
Tokens 17.4T
Requests 167.8M
Capabilities and verification
Reviewed against the HarnessRouter capability catalog onAugust 24, 2026. Every verified row cites its public source; rule-derived rows apply HarnessRouter's published classification rules to documented behavior. A “No public evidence” badge marks a gap in the public record, not a statement thatClaude Code lacks them.
Configuration & support
- Other models No public evidence
- Custom instructions Directly verified
- Skills Directly verified
- Custom tools Directly verified
- MCP Directly verified
- Sandbox & approvals Directly verified
- Sessions Directly verified
- Subagents Directly verified
Execution & agency
- Workspace files Directly verified
- Structured actions Directly verified
Environment control
- Browser control Directly verified
- Computer control Directly verified
Observation & capture
- Screen observation Directly verified
- Screenshot export No public evidence
- Recording export Directly verified
Artifact outputs
- Documents Rule-derived
- Spreadsheets Rule-derived
- Presentations Rule-derived
- Images Rule-derived
- Video Rule-derived
- App deployment No public evidence
Sources:code.claude.com
Machine interfaces
- Headless CLI ·Stable
- SDK ·Stable
Benchmark evidence
Claude Code configurations from the HarnessRouter Care Prep benchmark, publishedAugust 10, 2026. One controlled synthetic-data task; results apply to this test, not universally.
| Model | End-to-end latency | Cost |
|---|---|---|
| claude-sonnet-4.6 | 1m 31s | 83.5 credits |
| claude-opus-4.8 | 3m 18s | 223 credits |
Execution success and strict grounding pass rates for these configurations are published in the observed-history capture on the benchmark report.
Run Claude Code through HarnessRouter
API harness ID: claude-code
Use it for code changes or finished documents that your product can review.