skills.rest
mkalkere
mkalkereCOMMUNITY
Explore all skills by this author →
GitHub

repo-intelligence

Analyze repositories and store structured intelligence in agent memory.

4|Updated Feb 1, 2026
Search Tags:#onboarding#analysis#patterns#repository#memory#conventions#intelligence
One-click install
npx skills add https://github.com/mkalkere/agent-coordinator --skill repo-intelligence-mkalkere
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: repo-intelligence
Source: https://github.com/mkalkere/agent-coordinator/tree/main/.os/skills/core/repo-intelligence
Command: npx skills add https://github.com/mkalkere/agent-coordinator --skill repo-intelligence-mkalkere

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Repository context often decays across sessions, causing misaligned decisions and duplicated effort. This Skill builds and preserves a structured understanding of a codebase to enable informed work over time.

Core Features & Use Cases

  • Initial Analysis: Detect language, frameworks, patterns, and file relationships to bootstrap context.
  • Persistent Context: Store intelligence in agent memory (repo_context.yaml, file_relationships.yaml, known_issues.yaml, conventions.yaml, trajectories/) for reuse across sessions.
  • Query Before Work: Load relevant context before coding or reviewing to reduce context switching and improve accuracy.
  • Update After Work: Append new insights and trajectories after PRs or reviews to refine understanding.
  • Onboarding & Diagnosis: Use during onboarding to accelerate ramp-up and during problem diagnosis to surface known issues and patterns.

Quick Start

Run the repo-intelligence skill to analyze your repository and load persistent context into memory for future tasks.

Frequently Asked Questions about repo-intelligence

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

FAQPage Schema
How do I maintain repository context across sessions to avoid misaligned coding decisions?▼

Maintain repository context across sessions by automating the accumulation and retrieval of codebase intelligence. This Skill stores language stacks, conventions, and file relationships in agent memory to prevent misaligned coding decisions and duplicated effort.

What is the best way to load codebase conventions before starting a code review?▼

The best way to load codebase conventions before a code review is to query the stored repository intelligence. This Skill provides a builtin query tool to retrieve patterns, conventions, and review histories on demand.

How does repository intelligence analysis work for onboarding to a new codebase?▼

Repository intelligence analysis works for onboarding by detecting languages, frameworks, and file relationships to bootstrap context. It stores this structured understanding in memory to accelerate ramp-up and surface known issues.

Can I update repository knowledge automatically after a pull request?▼

Yes, you can update repository knowledge after a pull request. This Skill appends new insights and trajectories to agent memory, refining the structured understanding of conventions and known issues following PRs and reviews.

Does this repository intelligence tool require any external dependencies to store analysis data?▼

No external dependencies are required to store analysis data. This Skill relies on a built-in indexing model that saves structured intelligence into agent memory files like repo_context.yaml and known_issues.yaml.

When do I need to run a repository analysis to capture file relationships and known issues?▼

Run repository analysis before coding or reviewing to capture file relationships and known issues. It is also needed during problem diagnosis to surface relevant patterns and during onboarding to establish baseline codebase context.

Related Skills

Explore More Software Engineering Skills→
alex-voloshin-devalex-voloshin-dev
2

ai-skills-init

Detect codebase type and initialize configuration scaffolding and memory directories.

Community
Advanced
kid0317kid0317
26

insights

Store fragmented work reflections as searchable, categorized notes in memory archives.

Community
Intermediate
AvaTar-ArTsAvaTar-ArTs

requesting-code-review

Automate pre-merge code-review tasks by dispatching a review subagent.

Community
Intermediate
phuhao00phuhao00
12

code-analyst-agent

Analyze source code to identify symbols, call graphs, and architectural patterns.

Community
Intermediate