commit / research

Commit research drafts into durable artifacts with external_refs receipts.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/Collectively-Gary/GaryOS --skill commit-research-collectively-gary
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: commit / research
Source: https://github.com/Collectively-Gary/GaryOS/tree/main/data/skills/commit/research
Command: npx skills add https://github.com/Collectively-Gary/GaryOS --skill commit-research-collectively-gary

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill safely transitions a research draft into a durable commit artifact by verifying required files, publishing to the right external destination when configured, and recording proof via receipts and external_refs.

Core Features & Use Cases

  • Deterministic commit for research briefs: Confirms the declared brief path exists, then seals it either as a local durable file or as an externally published artifact with captured identifiers.
  • Variant-aware irreversible actions: Handles rejections, dataset, and calibration learning variants with correct artifact consumption, validation, and external filing/receipt behavior.
  • Guardrails and explicit failure gates: Escalates to awaiting_operator or awaiting_external for missing tools, missing credentials, malformed learning_variant values, inconsistent datasets, or missing/invalid draft artifacts.
  • Receipt integrity: Appends authoritative external_refs to index.md so verify can re-fetch and validate what was committed.

Quick Start

Ask an operator to commit a playbook: research action by setting learning_variant appropriately so the commit worker publishes (when configured) and records receipt external_refs for later verification.

Frequently Asked Questions about commit / research

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

FAQPage Schema
How do I automate research artifact publication and record external references?

To automate research artifact publication, verify the declared draft path exists, execute the irreversible publication step via configured tools, and append external_refs receipts to index.md for later verification.

What is the process for committing dataset validation checks in a research workflow?

Dataset validation checks in a research workflow require verifying dataset consistency and schema before committing artifacts. Failures trigger strict escalation to awaiting_operator or awaiting_external to prevent invalid commits.

How do I handle rejections and calibration variants when publishing research outputs?

Handling rejections and calibration variants requires variant-aware artifact consumption, validation, and external filing behavior. The workflow ensures correct receipt recording for each specific learning_variant during the commit stage.

Why does my GitHub issue filing fail during the research commit stage?

GitHub issue filing fails during the research commit stage when GH_TOKEN credentials are missing or invalid. The workflow enforces atomic preconditions and escalates to awaiting_external on missing tools or external publication failures.

What are the prerequisites for durable research artifact publishing?

Prerequisites for durable research artifact publishing include existing draft artifacts, valid learning_variant values, and configured external publication tools. The workflow enforces atomic preconditions and explicit failure gates before committing.

When should I escalate to awaiting_operator status in a research commit workflow?

Escalate to awaiting_operator status when encountering missing tools, missing credentials, malformed learning_variant values, inconsistent datasets, or missing draft artifacts. This strict escalation prevents invalid irreversible publication steps.