Skip to content

Agent Claude Harness

Component id: agent_claude

Source contract: BlockContract for agent_claude.

Overview

Beta, review-pending Anthropic Claude agent with tool calling, autonomous tool-use loops, retry policy, circuit breaker, and lifecycle management. Block contract verification_method=tier1_review_pending; @component_maturity("beta").

When to use:

  • Single-shot Claude inference with tool calling (infer())
  • Autonomous multi-step agent execution: call → observe → execute → repeat (infer_loop())
  • Production deployments requiring retry, circuit breaker, and structured logging

Good at

  • single-shot Claude inference with tool calling
  • autonomous multi-step tool-use loops (call → observe → execute → repeat)
  • deployments needing retry, circuit breaker, and structured logging

Not suited for

  • enforcing a central spend cap (paid calls are consent-gated, not quota-enforced)
  • running where the Claude Code CLI is absent from PATH or unauthenticated

Mitigates

Known Failure Modes

  • review_pending: operator_review

State Surface

  • Persistence: run
  • Queryable via: snapshot_export
  • Retention: run

Extension Points

  • training_signal_inputs: ('agent_routing_outcomes',)
  • validation_gate: unit_plus_codex_review
  • allowed_mutation_surfaces: ('prompts', 'routing_policy')

Verification

  • Method: tier1_review_pending
  • Human review gate: False
  • Fallback block: none