skills.rest
TeXmeijin
TeXmeijinCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

isis

Investigates issues and vague requests as hypotheses against codebase facts before implementation.

Updated May 20, 2026
Search Tags:#investigation#github issues#requirements analysis#hypothesis testing#linear#decision support#issue triage
One-click install
npx skills add https://github.com/TeXmeijin/agent-skills --skill isis-texmeijin
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: isis
Source: https://github.com/TeXmeijin/agent-skills/tree/main/.apm/skills/isis
Command: npx skills add https://github.com/TeXmeijin/agent-skills --skill isis-texmeijin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams often jump into implementation based on vague tickets, rough verbal requests, or unverified assumptions, wasting effort on the wrong problem. This Skill treats every Issue or request as a hypothesis and re-evaluates it against the actual codebase, data, and existing specs before any code is written. ## Core Features & Use Cases - Hypothesis-based issue triage: Fetches Linear issues (via MCP) or GitHub issues (via gh CLI), decomposes claims, and separates confirmed facts from unverified assumptions. - Fact-grounded investigation: Searches code, git history, logs, and monitoring data to validate whether the reported problem is real, its scope, and whether similar solutions already exist. - Structured decision support: Produces a re-evaluation report with options (do nothing, measure first, minimal fix, full implementation) plus closed-ended questions (A/B/C format) that minimize the requester's reply cost. - Use Case: A teammate pastes "TES-123調べて" into chat. The Skill fetches the Linear issue, checks the relevant code and git history, finds the reported bug is actually intended behavior, and presents a report recommending monitoring instead of a fix. ## Quick Start Ask the agent to investigate and re-evaluate a specific Linear or GitHub issue before starting implementation, for example by saying "このIssueについて調査して" with the issue ID.

Frequently Asked Questions about isis

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

FAQPage Schema
How do I investigate a Linear or GitHub issue before implementing it?▼

Pass the Linear issue ID (e.g., TES-123) or GitHub issue URL to the Skill. It fetches the issue via the Linear MCP server or gh CLI, decomposes the claims, verifies them against the codebase and git history, and outputs a structured re-evaluation report.

How to turn a vague feature request into an actionable task definition?▼

Submit the rough request as text. The Skill treats it as a hypothesis, identifies ambiguities and hidden assumptions, checks facts in the code, then produces a goal definition with closed-ended A/B/C questions so the requester can confirm scope with minimal effort.

Does this Skill start implementing code after analyzing an issue?▼

No. Implementation without agreement is explicitly forbidden. It stops at a recommendation report with options and a minimal plan, and is designed to run before Plan Mode, which handles the actual design and implementation phase.

What happens when the requester does not answer the clarification questions?▼

The Skill applies default provisional assumptions: minimal scope, investigation-first deliverable, and backward compatibility preserved. It separates items that are safe to proceed with from those that are dangerous without an answer, rather than blocking entirely.

When should I not use hypothesis-based issue investigation?▼

Skip it when the request is already well-specified with clear success criteria; the Skill itself detects this and only lightly reformats the request. It is also unnecessary for trivial mechanical changes with no ambiguity about the problem being solved.

Related Skills

Explore More Product & Management Skills→
nesihaver-ILnesihaver-IL

create-issue

Create structured issues in GitHub Issues, Linear, or Jira.

Community
Intermediate
cornmanwtfcornmanwtf

chaos-engineering-runner

Plan fault injection and resilience tests with scope, constraints, and success metrics.

Community
Advanced
WhaleylawWhaleylaw

error-tracking-specialist

Analyze stack traces and error context to triage production application errors.

Community
Intermediate
afoxnyc3afoxnyc3

cost-review

Analyze Azure Cost Management data for spend drivers and budget variance.

Community
Intermediate