extract-evaluation-protocol

SkillDev tools

Lets your agent pull evaluation details like metrics and baselines from a source and flag what's missing or unclear.

Available today. Use it from your connected AI after setup.

Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.

Then ask your AI: use the extract-evaluation-protocol skill

About this skill

Extract evaluation-protocol elements from a source and mark unstated or ambiguous parameters.

What this skill tells your AI

The instructions your AI receives, as published by yogsoth-ai/de-anthropocentric-research-engine in skills/extract-evaluation-protocol/SKILL.md and read by ahel’s review.

Purpose

Extract evaluation-protocol elements from a source and mark unstated or ambiguous parameters.

Input contract

required: [source_record, protocol_schema]
optional: [supplementary_materials, domain_defaults]
constraints: [extracted values must distinguish stated, inferred, and missing]

Procedure

  1. Parse task, data, split, metric, baseline, budget, evaluator, and reporting elements.
  2. Link each value to a source passage or record field.
  3. Mark ambiguity, omission, and any inference made from context.
  4. Emit a normalized protocol record with comparability notes.

If multiple protocols can now be represented on the same comparison schema, consider compare-evaluation-protocols as the next tactic.

Output contract

produces: [protocol_record, source_links, ambiguity_log, comparability_notes]
delta_fields: [evidence_updates, uncertainties, open_questions]

Quality gates

  • Missing fields are explicit.
  • Inferred defaults are never represented as stated facts.

Failure and counterexamples

Do not fill omitted protocol values from common practice without an inference label.

Provenance map

  • resolved: protocol-element-extraction

Signals

GitHub stars
501
Forks
41
Last commit
Sep 2026
Advanced
Catalog kind
skill
Key
extract-evaluation-protocol
Source
github.com/yogsoth-ai/de-anthropocentric-research-engine