Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/LIDR-academy/lidr-specboot/llms.txt

Use this file to discover all available pages before exploring further.

Skills are SKILL.md files stored under ai-specs/skills/ — self-contained reference guides that encode proven techniques, patterns, and workflows so any AI agent can apply them consistently across sessions and projects. Rather than rediscovering the right approach each time, agents load the relevant skill, follow its instructions, and produce repeatable results.

​
How Auto-Loading Works

Agents read base-standards.md Section 4, which instructs them to scan the skills directory and automatically load any skill whose description field matches the current request. This means you rarely need to invoke a skill by name; the agent recognises the situation and applies the correct workflow on its own. Skills are also available for explicit invocation using the /skill-name command syntax supported by OpenSpec. Skills live canonically in ai-specs/skills/. To make them discoverable by Claude and Cursor without duplicating files, Specboot creates relative symlinks:
  • .claude/skills/ → entries in ai-specs/skills/
  • .cursor/skills/ → entries in ai-specs/skills/
This means updates to a skill file in ai-specs/skills/ are immediately reflected everywhere. After any rename, move, or new addition, use the sync-agent-symlinks skill to keep the symlink tree consistent.
The sync-agent-symlinks skill automates symlink maintenance. Run it after adding a new skill or renaming an existing one to ensure .claude/skills/ and .cursor/skills/ stay in sync with ai-specs/skills/.

​
All Included Skills

SkillDescription
enrich-usAnalyze and enhance vague user stories into implementation-ready tickets with acceptance criteria, technical detail, and edge cases
code-auditingRun a systematic 7-phase code quality audit (Phase 0–6) covering security, performance, type safety, dead code, and library best practices
commitStage relevant changes, write descriptive commit messages, push the branch, and create or update a GitHub PR
using-git-worktreesSet up an isolated git worktree before feature work, with baseline checks, Claude settings propagation, and safe cleanup
writing-skillsAuthor and verify new skills following TDD-style RED-GREEN-REFACTOR validation before deployment
adversarial-reviewAct as an independent red-team reviewer after implementation, before archiving an OpenSpec change
update-docsIdentify and update required technical documentation based on implemented changes
explainTeach underlying concepts with clear mental models to close the skill gap behind a user’s question
meta-promptRewrite prompts using prompt-engineering best practices for precise and complete results
openspec-sync-specsSynchronise OpenSpec artifacts with the current implementation state
show-spec-workingDisplay the specification context being applied to an active change or session
sync-agent-symlinksRecreate .claude/skills/ and .cursor/skills/ symlinks to match the current ai-specs/skills/ directory

​
Documented Skills

enrich-us

Transform vague user stories into implementation-ready tickets with acceptance criteria and technical detail.

code-auditing

Systematic 7-phase (Phase 0–6) code quality audit covering security, performance, type safety, and dead code.

commit

Create focused commits with descriptive messages, push branches, and manage GitHub PRs via the gh CLI.

Git Worktrees

Set up isolated workspaces for feature development with safe creation, baseline verification, and cleanup.

writing-skills

Apply Test-Driven Development to skill authoring: write scenarios, watch agents fail, write the skill, verify compliance.

​
Skill File Structure

Every skill follows a standard layout that makes it machine-readable and human-friendly: 
---
name: skill-name
description: Use when [specific triggering conditions and symptoms]
author: LIDR.co
version: 1.0.0
---

# Skill Name

## Overview
Core principle in 1-2 sentences.

## When to Use
Bullet list of symptoms and situations.

## Instructions
Step-by-step process with code examples.

## Notes
Edge cases and important caveats.
The YAML frontmatter’s description field is what the agent reads to decide whether to load the skill. It should start with “Use when…” and describe triggering conditions only — never summarise the workflow itself, as that causes agents to follow the description instead of reading the full skill body.

​
Creating a New Skill

  1. Create the directory and main file: ai-specs/skills/{name}/SKILL.md
  2. Write YAML frontmatter with name, description, author, and version
  3. Follow the writing-skills workflow (TDD: run a baseline scenario first, write the skill, verify compliance, close loopholes)
  4. Add any supporting reference files inside the same directory if needed
  5. Run sync-agent-symlinks to create the corresponding symlinks in .claude/skills/ and .cursor/skills/
For a detailed walkthrough of the authoring process, see the writing-skills skill documentation.

Build docs developers (and LLMs) love

Get started for freeTalk to us