Measured harness ledgerPublic result
Claude Fable 5.1Neon Gunner — Claude Fable 5.1 Max
Build a playable destructible neon run-and-gun game in the browser under the pinned Harness task contract.
Max reasoningHeadline result
- Workflow cost
- $95.09
- Wall-clock
- 55m 21.3s wall-clock
- Processed tokens
- 44.65M processed
- Record state
- partial_active_session_snapshot_with_supplied_validation
Public summary
Claude Fable 5.1 Max partial_active_session_snapshot_with_supplied_validation ledger: 55m 21.3s wall-clock, 44.65M processed, and $95.09 Official Anthropic API-list-price equivalent from the supplied active-session snapshot, not an itemized Claude Code subscription charge..
Run identity and stack
- Result ID: neon-gunner-canvas-fable-5.1-max
- Technical model: claude-fable-5-1
- Provider: Anthropic Claude Code
- Stack: Anthropic Claude Code
- Stack: Technical model/configuration: claude-fable-5-1
- Stack: Browser canvas game
- Stack: Harness v1 Neon Gunner prompt
Cost basis
- The supplied receipt reports that it fetched and verified the listed rate card from the official Anthropic pricing page on 2026-09-02; the archive preserves that provenance claim.
Primary artifact integrity
- Kind: manifest-verified-artifact
- Path: artifacts/neon-gunner-canvas-fable-5.1-max/source/index.html
- SHA-256: f3ba201c1f48c60bfda93037e34d709a01086f1808cc5b0f8fb200a3a829e90e
Validation evidence
- Result: PARTIAL_PASS
- Path: artifacts/neon-gunner-canvas-fable-5.1-max/validation/archive-verification.public.json
- SHA-256: 5d978ac8a2734b4692f2a508f55df5bf820317ab9170d43aca0149b9b4269abe
- Validator SHA-256: fa5ec13e9ef2f5f96949283077b7c6278bf9c9f14032817c21743ab973d06090
Recorded caveats
- The metrics are a snapshot while the source Claude Code session remained active; final transcript totals may be slightly larger.
- Wall-clock is end-to-end workflow latency, including tool execution, network, and waiting; it is not model-only compute time.
- Output tokens include hidden reasoning, generated code, and tool-call JSON, not only visible prose.
- Cache-read tokens are billed at a deep discount, so total processed tokens materially overstate effective cost.
- The 26/26 validator log is supplied model-run evidence. The archival smoke check independently exercised only title entry and a jump with real browser input; it did not traverse the level, exercise every combat verb, or replace human playtesting.
- No local FPS, browser hardware, standardized final capture, or blind-evaluation receipt is supplied.
Visible evidence gaps
- final inactive-session token and cost receipt
- independent full human-playthrough record
- local FPS, browser, viewport, and hardware receipt
- standardized final capture
- blind-evaluation record
Public result only
This result keeps its public summary and evidence, but it does not currently have a matching Builder test with prompts, projects, Harness workflows or skills.