Capture a claim
- Read
rigorgraph.yaml, current claims, evidence, and verifications.
- Reply in the explicit or configured language: English, Traditional Chinese, Simplified Chinese, or Japanese. Keep claim statements, quotations, formulas, identifiers, and source titles in their original language.
- Classify the claim as
formal, literature, empirical, benchmark, or synthesis.
- State the exact scope. For a source, record its URI and precise locator. For a local artifact, record a project-relative path and SHA-256 digest.
- Classify each evidence record as
proof, source, dataset, computation, benchmark_run, or human_review. Never relabel numerical evidence as proof.
- Create evidence JSON first and run
rigorgraph evidence add. Then create a DRAFT or PROPOSED claim JSON and run rigorgraph claim add.
- When the input is a versioned RigorGraph evidence bundle, preserve it with
rigorgraph evidence import BUNDLE.json --claim CLAIM_ID instead of rewriting producer metadata. Bundle hashes record observed bytes; they do not prove producer identity, runner authenticity, or correctness.
- Run
rigorgraph audit. If it fails, report the exact issue codes and leave the epistemic status unchanged.
Use localized result headings:
| en |
zh-TW |
zh-CN |
ja |
| Claim recorded |
已記錄命題 |
已记录命题 |
主張を記録しました |
| Evidence class |
證據類別 |
证据类别 |
証拠区分 |
| Supported scope |
支持範圍 |
支持范围 |
支持される範囲 |
| Remaining gap |
剩餘缺口 |
剩余缺口 |
残るギャップ |
Do not invent a locator, digest, author, verifier, experimental condition, or dependency.
1---2name: capture-claim3description: Capture a research claim and its scoped supporting evidence in a RigorGraph project. Use when recording a theorem target, literature result, empirical observation, benchmark result, synthesis, source, dataset, computation, proof packet, or human review.4---56# Capture a claim781. Read `rigorgraph.yaml`, current claims, evidence, and verifications.92. Reply in the explicit or configured language: English, Traditional Chinese, Simplified Chinese, or Japanese. Keep claim statements, quotations, formulas, identifiers, and source titles in their original language.103. Classify the claim as `formal`, `literature`, `empirical`, `benchmark`, or `synthesis`.114. State the exact scope. For a source, record its URI and precise locator. For a local artifact, record a project-relative path and SHA-256 digest.125. Classify each evidence record as `proof`, `source`, `dataset`, `computation`, `benchmark_run`, or `human_review`. Never relabel numerical evidence as proof.136. Create evidence JSON first and run `rigorgraph evidence add`. Then create a `DRAFT` or `PROPOSED` claim JSON and run `rigorgraph claim add`.147. When the input is a versioned RigorGraph evidence bundle, preserve it with `rigorgraph evidence import BUNDLE.json --claim CLAIM_ID` instead of rewriting producer metadata. Bundle hashes record observed bytes; they do not prove producer identity, runner authenticity, or correctness.158. Run `rigorgraph audit`. If it fails, report the exact issue codes and leave the epistemic status unchanged.1617Use localized result headings:1819| en | zh-TW | zh-CN | ja |20| --- | --- | --- | --- |21| Claim recorded | 已記錄命題 | 已记录命题 | 主張を記録しました |22| Evidence class | 證據類別 | 证据类别 | 証拠区分 |23| Supported scope | 支持範圍 | 支持范围 | 支持される範囲 |24| Remaining gap | 剩餘缺口 | 剩余缺口 | 残るギャップ |2526Do not invent a locator, digest, author, verifier, experimental condition, or dependency.
f0909172434