corgispec-ask

Answer pending Obsidian questions with early-stop retrieval under a token budget.

113|9|Updated Apr 30, 2026
One-click install
npx skills add https://github.com/ricoyudog/Coding_Corgi_flow --skill corgispec-ask
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: corgispec-ask
Source: https://github.com/ricoyudog/Coding_Corgi_flow/tree/main/.opencode/skills/molecules/corgispec-ask
Command: npx skills add https://github.com/ricoyudog/Coding_Corgi_flow --skill corgispec-ask

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents inefficient, runaway Q&A over an Obsidian vault by enforcing a strict retrieval budget and stopping as soon as enough information is found to answer the question.

Core Features & Use Cases

  • Early-stop retrieval: Searches the vault using a prioritized context chain and stops once the answer can be produced confidently.
  • Token/budget guardrails: Limits the number of wiki/domain pages read and total file reads to avoid excessive context usage.
  • Question lifecycle writeback: Updates question frontmatter status and writes the answer, sources, and (when needed) a needs-deep-session plan back to the same markdown file.
  • When to use: Use it when humans create wiki/questions/*.md files with status: pending, then process them via a single-file command or batch --pending.

Quick Start

Run the command to answer a specific pending question file: /corgi-ask wiki/questions/<filename>.md

Frequently Asked Questions about corgispec-ask

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

FAQPage Schema
How do I answer pending questions in an Obsidian vault without exceeding token limits?

Budgeted retrieval answers Obsidian vault questions by enforcing a strict token budget. It searches memory and wiki indices using a prioritized context chain, stopping early once enough information is found, then writes the answer and sources back to the markdown file.

What is early-stop retrieval for question answering workflows?

Early-stop retrieval is a question answering mechanism that searches a prioritized context chain across memory and wiki indices, halting file reads as soon as it confidently produces an answer to prevent runaway context consumption.

How to process batch pending questions from a wiki directory?

Process batch pending questions by running a batch command targeting the wiki/questions/ directory. The workflow reads markdown files with pending status, retrieves context within token guardrails, and writes back the answer, sources, and status updates to each file.

Does question answering with budgeted retrieval write answers back to the source markdown files?

Yes, budgeted retrieval writes back to source markdown files. It updates frontmatter status and records the answer, retrieved sources, and when needed a needs-deep-session plan directly into the original question file.

What are the limitations of using a strict token budget for knowledge retrieval?

Limitations of strict token budget retrieval include capping the number of wiki and domain pages read, which may leave complex queries unanswered. It flags such files with a needs-deep-session plan instead of forcing a partial answer.