openspec-explore

Investigates 1C defects and questions, producing a structured change-request brief in chat.

Updated Jun 26, 2026
One-click install
npx skills add https://github.com/pavelvdo/universal-xml-exchange2 --skill openspec-explore-pavelvdo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: openspec-explore
Source: https://github.com/pavelvdo/universal-xml-exchange2/tree/main/.cursor/skills/openspec-explore
Command: npx skills add https://github.com/pavelvdo/universal-xml-exchange2 --skill openspec-explore-pavelvdo

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? When a 1C developer receives a raw bug report, trace file, or vague customer question, this Skill turns it into a structured investigation: it triages the input, delegates trace and code analysis to specialized agents, and ends with a ready-to-use change request (ZNi) statement instead of an unstructured chat log. ## Core Features & Use Cases - Unified entry point: Any free-form text (bug, question, raw customer text, screenshots) enters one explore flow with a mandatory brief card (symptom, route, question, confirmation) before any code reading. - Profile-based investigation: Bug profile analyzes traces (.pff, TRACE) via a trace-analyst agent; question profile answers how-things-work queries via a code-explorer agent; feature requests redirect to /opsx:new. - Structured handoff: Bug investigations finish with a ## Постановка ЗНИ block in chat (symptom, verified root cause, fix approach, acceptance criteria) that downstream commands /opsx:new and /opsx:extend consume directly. - Use Case: A customer reports "valid Kontur signature shows as invalid" with a trace file. The Skill builds a brief, runs trace analysis, checks archive relevance, and outputs a complete change-request statement with verified root cause and acceptance criteria. ## Quick Start Type /opsx:explore followed by a description of your 1C defect, question, or paste the raw customer text to receive a structured brief and investigation plan.

Frequently Asked Questions about openspec-explore

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

FAQPage Schema
How do I investigate a 1C bug with a trace file?

Run /opsx:explore with the symptom description and trace path. The Skill builds a brief card for confirmation, then delegates trace analysis to a trace-analyst agent whose full report is saved to temp/reports/ with a chat preview and verdict.

What is the difference between explore-bug and explore-question profiles?

The bug profile handles defects, errors, and traces, ending with a formal change-request statement block. The question profile answers how-things-work queries with a chat summary and does not propose a change request by default.

Does openspec-explore modify 1C source code (.bsl files)?

No. The Skill is investigation-only and explicitly forbidden from writing or editing .bsl files. Code changes happen later through the separate /opsx:apply command after a change request is created via /opsx:new.

What are the prerequisites for using this Skill?

It requires the openspec CLI installed and an OpenSpec-initialized 1C project with openspec/project.md. The Skill runs inside Cursor with access to specialized agents like onec-trace-analyst and onec-code-explorer.

Why does the Skill refuse to give a 'works as designed' verdict?

A Symptom Lock rule forbids concluding 'expected behavior' while the user's original question remains unresolved. If findings do not explain the symptom, the Skill must name another cause or propose a concrete next investigation step.

When should I use /opsx:new instead of /opsx:explore?

Use /opsx:new directly when the request is a clear feature with obvious scope ("add X", "change behavior Y"). Explore is for unclear defects, questions, or raw customer input that needs investigation before a change can be specified.