skills.rest
truongnat
truongnatCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

init

Generates project reference documentation and merges repository-evidenced rules into agent settings.

Updated Jul 28, 2026
Search Tags:#monorepo#project context#documentation generation#agent configuration#repository scanning#stack detection
One-click install
npx skills add https://github.com/truongnat/Restly --skill init-truongnat
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: init
Source: https://github.com/truongnat/Restly/tree/main/.agents/skills/init
Command: npx skills add https://github.com/truongnat/Restly --skill init-truongnat

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? AI agents lack reliable project context when entering a repository, leading to guessed commands, invented conventions, and missed stacks in multi-ecosystem monorepos. This Skill builds a sourced, confidence-classified project reference so every subsequent agent decision is grounded in evidence. ## Core Features & Use Cases - Read-only repository discovery: Inventories git metadata, manifests, configs, CI, tests, and docs without executing project code or reading secrets. - Deep workspace scanning: Runs a bundled filesystem sweep script to surface every project root, then classifies each stack (Node, Flutter, Go, Rust, Python, and more) so monorepo stacks are never collapsed into one. - Structured artifacts: Produces .agents/PRJ_REFERENCE.md from a template with executive summary, tech stack, business rules, commands, and unknowns, plus a merged .agents/settings.yaml that preserves user values. - Use Case: An agent joins a monorepo containing a Next.js app and a Flutter app. Running this Skill scans the filesystem, records both stacks separately with their own commands, and flags undocumented conventions as unknowns for user confirmation. ## Quick Start Ask the agent to initialize or refresh the project reference and settings for this repository using the init skill.

Frequently Asked Questions about init

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

FAQPage Schema
How do I give an AI agent context about my repository?▼

Run the init skill to scan the repository and generate a PRJ_REFERENCE.md containing the tech stack, architecture, commands, and conventions with sources. Subsequent skills read this file before making project-specific decisions.

How do I detect all stacks in a monorepo?▼

Run the bundled scan_workspaces.sh script, which sweeps the filesystem rather than the workspace config to surface every project root. The agent then classifies each candidate's stack from its actual manifest, covering ecosystems like Flutter, Go, Rust, and Python.

Does the init skill modify my source code or settings?▼

Discovery is read-only and never executes project code or mutates source files. It merges only repository-evidenced or user-confirmed rules into settings.yaml while preserving language and custom user values.

Can repository scanning expose secrets or credentials?▼

No. The skill forbids opening .env files, credential stores, private keys, or secret-manager payloads. Git remotes are recorded with credentials redacted, and agent CLI detection never stores tokens.

What is the difference between init, refresh, and force modes?▼

Init creates the reference from a template when none exists, refresh updates changed facts while preserving confirmed content, and force rebuilds everything from a full re-scan when the user explicitly requests regeneration.

Related Skills

Explore More Software Engineering Skills→
dlwlrmaxdlwlrmax

clonedeps

Clone dependency source repositories into an ignored local workspace for agent inspection.

Community
Intermediate
AkbarDevopAkbarDevop

autoplan

Run CEO, design, and eng review skills sequentially with decision principles.

Community
Advanced
AjayRajan05AjayRajan05
5

github-auth

Configures GitHub authentication using HTTPS tokens, SSH keys, or the gh CLI.

Community
Intermediate
jiten-singh-shahijiten-singh-shahi
13

security-scan

Audit Claude Code configuration files for security vulnerabilities and emit JSON findings.

Community
Intermediate