evidence-protocol-module

Enforce citation-backed evidence documentation before releasing final answers.

43|2|Updated Mar 26, 2026
One-click install
npx skills add https://github.com/curated-skills/LinguaClaw --skill evidence-protocol-module
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: evidence-protocol-module
Source: https://github.com/curated-skills/LinguaClaw/tree/main/harnesses/modules/evidence-protocol
Command: npx skills add https://github.com/curated-skills/LinguaClaw --skill evidence-protocol-module

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents premature releases by enforcing that the agent writes a standalone, citation-backed evidence document before any final answer is allowed to appear.

Core Features & Use Cases

  • Standalone evidence artifact: Requires an evidence document (defaulting to analysis.md in the stage artifact directory) to exist and be materially complete before releasing results.
  • Structured, reviewable reasoning: Demands specific sections covering problem framing, relevant components/files, observed failures, root-cause causal chains, proposed changes with claim-by-claim mapping, validation outcomes, and residual uncertainty.
  • Claim discipline with provenance: Instructs that every important claim must state what it is, what observation/output/source anchors it, whether it is observed vs inferred, and what uncertainty remains.
  • Answer gated release policy: Blocks final answer release when release-critical claims are uncited, logically inconsistent, or unresolved.

Quick Start

Use the evidence-protocol module to require a standalone evidence document (analysis.md) and block any final answer until it includes complete, claim-scoped citations that satisfy the protocol.

Frequently Asked Questions about evidence-protocol-module

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I enforce citation-backed evidence documentation before releasing agent answers?

To enforce citation discipline, you require a standalone evidence artifact with structured sections covering problem framing, root-cause chains, and claim provenance before an answer release gating policy permits final output.

What is release gating for debugging and root cause analysis?

Release gating for root cause analysis is a blocking policy that prevents final answers from appearing when release-critical claims are uncited, logically inconsistent, or contain unresolved contradictions within the evidence documentation.

How do I structure an evidence document for experimental validation and patch justification?

Structure the evidence document with specific sections covering problem framing, relevant components, observed failures, root-cause causal chains, proposed changes with claim mapping, validation outcomes, and residual uncertainty.

How does claim provenance discipline work for agent harness outputs?

Claim provenance discipline works by requiring every important claim to explicitly state what it is, what observation or source anchors it, whether it is observed versus inferred, and what residual uncertainty remains.

What happens if critical evidence gaps remain during debugging analysis?

If critical evidence gaps remain, the release gating policy blocks the final answer release until the unresolved contradictions or uncited release-critical claims are materially complete and satisfy the evidence protocol.