spectra-ask

Retrieve specification knowledge from openspec/ documents via semantic vector search.

Updated Apr 27, 2026
One-click install
npx skills add https://github.com/zooroy/appt-system --skill spectra-ask-zooroy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: spectra-ask
Source: https://github.com/zooroy/appt-system/tree/main/.agents/skills/spectra-ask
Command: npx skills add https://github.com/zooroy/appt-system --skill spectra-ask-zooroy

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires spectra CLI.

What problem does it solve?

It helps you quickly find and answer questions by searching the project’s specification documents under openspec and grounding the response in what the files actually say.

Core Features & Use Cases

  • Spec-grounded Q&A: Answers are required to be based only on documents under openspec, with no reliance on outside knowledge.
  • Semantic search over specs: Uses embedding-based vector search to match cross-language queries (e.g., Chinese/English/Japanese) to relevant spec content.
  • Operational steps for retrieval: Runs spectra search, reads matched files (up to 10), and prioritizes openspec/specs over openspec/changes/archive when topics overlap.
  • When content is missing: If the documents don’t contain the answer, it responds that the spec files don’t have the needed content.

Quick Start

Ask a question like: $spectra-ask how the availability calculation handles past time slots?

Frequently Asked Questions about spectra-ask

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

FAQPage Schema
How do I get answers from openspec specification documents?

To get answers from openspec specification documents, you ask a question and the system runs a semantic vector search over the repository, reading up to 10 matched files to generate a grounded response based strictly on the retrieved spec content.

How does semantic search work for project-feature discovery in spec files?

Semantic search for feature discovery uses embedding-based vector search to match cross-language queries to relevant spec content, prioritizing the openspec/specs directory over the openspec/changes/archive when overlapping topics are found.

Can I ask specification questions in languages other than English?

Yes, you can ask specification questions in different languages. The semantic vector search applies embedding-based matching to support cross-language queries, allowing users to retrieve relevant openspec document content in Chinese, English, or Japanese.

What happens when the spec files don't contain an answer to my question?

When spec files don't contain the answer, the system explicitly responds that the specification documents lack the needed content, ensuring responses rely only on retrieved openspec files without outside knowledge.

Do I need the spectra CLI to search my knowledge base?

Yes, you need the spectra CLI to search your knowledge base. The question-answering process runs the spectra search command to find matched specification files and retrieve their content for grounded responses.