agent-skills-library
🥈SilverProduction-focused library of agent skills for programming workflows including code review, debugging, verification, and subagent orchestration.
1910Updated 2mo ago
Intermediate30min to implementautomation
Saves ~70 min per use
Quick InstallView Source
git clone https://github.com/CodingCossack/agent-skills-library.gitWorks with:
Claude
Overview
About This Skill
Agent Skills Library is a curated collection of skills designed for programming workflows with Claude and Cursor agents. It provides structured templates and executable steps for code review, systematic debugging, verification-before-completion, and subagent orchestration. Built as an improved rework of obra/superpowers, this library features tighter workflows, clearer templates, and enhanced guardrails. Each skill is production-focused with stronger execution order and verification gates, reducing meta-instructions while improving practical usability for development teams.
How to Use
[{"step":"Define Your Workflow Scope","action":"Identify the specific task to automate (e.g., code review, bug triage) and gather inputs: repository link, team roles, and existing tools (GitHub, Jenkins, Slack). Use Sortd’s [Use Case Templates](https://www.sortd.com/template-boards) for pre-built board structures.","tip":"Start with a single column (e.g., \"Code Review\") and expand to other columns (e.g., \"QA Failures\") as you refine the automation rules."},{"step":"Configure Automation Triggers","action":"Set up webhooks or APIs for your tools (GitHub, Jenkins, etc.) to push data into Sortd. In Sortd, go to **Settings > Integrations** and enable the relevant connectors. For GitHub, use the [Sortd GitHub App](https://github.com/marketplace/sortd).","tip":"Test triggers with a dummy event (e.g., create a test GitHub issue) to verify Sortd task creation before scaling."},{"step":"Design Your Sortd Board","action":"Create a new board in Sortd and customize columns based on your workflow. Use [AGENT_SKILLS_LIBRARY]’s templates for columns like \"Code Review\" or \"Debugging Queue\". Add labels (e.g., `critical`, `bug`) and automation rules (e.g., \"Move to QA Failures if Jenkins build fails 3 times\").","tip":"Use Sortd’s \"Automation Rules\" feature to auto-assign tasks based on file paths (e.g., frontend files → `@frontend-team`)."},{"step":"Deploy Subagents for Orchestration","action":"Integrate [AGENT_SKILLS_LIBRARY] subagents (e.g., `code-review-agent`, `debugging-agent`) with Sortd. Use Sortd’s API or webhook system to trigger agents when tasks move between columns. For example, when a task enters \"Code Review\", trigger the `code-review-agent` to analyze the PR.","tip":"Start with a single subagent (e.g., `code-review-agent`) and expand to others (e.g., `debugging-agent`) once the workflow stabilizes."},{"step":"Monitor and Iterate","action":"Use Sortd’s dashboard to track bottlenecks (e.g., tasks stuck in \"Code Review\"). Run weekly audits to refine automation rules and agent outputs. Adjust thresholds (e.g., Jenkins failure count) based on team feedback.","tip":"Enable Sortd’s \"AI Features\" (e.g., urgency detection) to auto-prioritize tasks and reduce manual sorting."}]
Use Cases
Automated code review and verification workflows
Systematic debugging and error diagnosis
Multi-agent task orchestration and planning
Pre-completion verification and quality gates
Tags
Setup & Installation
Quick Install
No install command available. Check the GitHub repository for manual installation instructions.
Alternative Install (Git Clone)
git clone https://github.com/CodingCossack/agent-skills-libraryRequirements
- Claude Code or compatible AI agent
- Works with: Claude
Quick Start Guide
1
Install the Skill
Copy the install command above and run it in your terminal.
2
Open Your AI Agent
Launch Claude Code, Cursor, or your preferred AI coding agent.
3
Try It Out
Use the prompt template or examples below to test the skill.
4
Customize
Adapt the skill to your specific use case and workflow.
Usage Examples
Prompt Template
Act as a Sortd (ai-assistant) automation agent integrated with [AGENT_SKILLS_LIBRARY]. Review the [PROJECT_NAME] repository in [REPOSITORY_LINK] and generate a prioritized action plan to automate [SPECIFIC_TASK] using Sortd’s kanban-style boards. Include: 1) Key automation triggers (e.g., GitHub webhooks, email notifications), 2) Sortd board setup instructions (columns, labels, automation rules), 3) Subagent orchestration steps for code review/debugging, and 4) Verification checkpoints. Reference [AGENT_SKILLS_LIBRARY] for code review templates and debugging workflows.
Example Output
### **Sortd Automation Plan for [PROJECT_NAME] Repository**
**Project Context:**
The [PROJECT_NAME] repository (https://github.com/company/project-name) is a Python-based CRM integration tool with 4 active contributors. Current pain points include: 1) Manual triage of GitHub issues into development tasks, 2) Lack of automated follow-ups for stale pull requests, and 3) No centralized tracking of code review feedback.
**Automation Triggers Identified:**
1. **GitHub Webhooks:** New issues labeled `bug` or `enhancement` trigger Sortd board creation in the "Triage" column.
2. **Email Notifications:** Jenkins build failures (e.g., `Build #42 failed`) auto-create Sortd tasks in "QA Failures" with Jira ticket links.
3. **Slack Alerts:** Critical PR reviews (e.g., `PR #123 requires changes`) generate Sortd tasks in "Code Review" with GitHub comment previews.
**Sortd Board Configuration:**
- **Board Name:** "[PROJECT_NAME] - DevOps Automation"
- **Columns:**
- **Triage** (Auto-populated via GitHub webhooks; labels: `bug`, `enhancement`, `documentation`)
- **Code Review** (Subagent: `code-review-agent` from [AGENT_SKILLS_LIBRARY]; auto-assigns to senior devs based on file paths)
- **QA Failures** (Auto-linked to Jenkins reports; severity labels: `critical`, `high`, `low`)
- **Ready for Merge** (Auto-archived after PR approval; Slack notification sent)
- **Done** (Auto-moved after 7 days of inactivity)
- **Automation Rules:**
- **Rule 1:** If a GitHub issue is labeled `bug`, create a Sortd task with the issue body as the description and assign to `@dev-team`.
- **Rule 2:** If a PR review comment contains `changes requested`, move the task to "Code Review" and notify `@reviewers` via Slack.
- **Rule 3:** If Jenkins build fails 3 times, escalate to "QA Failures" and tag `@qa-lead`.
**Subagent Orchestration:**
1. **Code Review Agent:**
- Trigger: New PR opened in `main` or `dev` branch.
- Action: Run `code-review-agent` from [AGENT_SKILLS_LIBRARY] to analyze diffs against [STYLE_GUIDE_LINK]. Output: Sortd task with:
- **Checklist:** [ ] PEP-8 compliance, [ ] Test coverage >80%, [ ] No hardcoded secrets.
- **Suggested Fixes:** Auto-generated inline comments for critical issues.
- Escalation: If `critical` severity, ping `@team-lead` in Sortd task.
2. **Debugging Agent:**
- Trigger: Jenkins failure with stack trace.
- Action: `debugging-agent` analyzes logs and suggests fixes. Output: Sortd task with:
- **Root Cause:** `TypeError: unsupported operand type(s) for +: 'int' and 'str'`
- **Fix:** Replace `+` with `str()` in `utils/parse_data.py` line 42.
- **Test Command:** `pytest tests/test_parse_data.py -v`
**Verification Checkpoints:**
- **Daily:** `code-review-agent` generates a Sortd report of all PRs older than 48 hours (auto-assigned to `@reviewers`).
- **Weekly:** `debugging-agent` runs a regression test on the top 5 most failed Jenkins builds and updates Sortd tasks with new insights.
- **Monthly:** Audit Sortd board for tasks stuck in "Code Review" >7 days and auto-escalate to `@team-lead`.
**Expected Outcome:**
- **Time Saved:** 4 hours/week on manual triage (based on 20 GitHub issues/month).
- **Quality Improvement:** 30% reduction in escaped bugs (projected from 5/month to <2/month).
- **Team Visibility:** Real-time Sortd dashboard shows bottlenecks (e.g., "Code Review" column avg. 3.2 days/issue).
**Next Steps:**
1. Set up GitHub webhook for Sortd (see [Sortd GitHub Integration Guide](https://docs.sortd.com/github)).
2. Configure Jenkins plugin to push build results to Sortd (use [Jenkins-Sortd API Key](https://api.sortd.com/jenkins)).
3. Deploy `code-review-agent` and `debugging-agent` via [AGENT_SKILLS_LIBRARY]’s subagent orchestration.
4. Train team on Sortd board columns and automation rules (15-minute onboarding video).Apply to these tools
Browse all toolsClaude
AI assistant built for thoughtful, nuanced conversation
Microsoft Teams
Get more done every day with Microsoft Teams – powered by AI
Drata
Automate security compliance and monitor real-time security posture seamlessly.
GPT for work
Automate your spreadsheet tasks with AI power
Respell
Agentic AI Workflow platform
Notion
Connected workspace for docs, wikis, and projects
Compatible MCP servers
Browse all MCP servers
Find the right skills for your stack
Take a free 3-minute scan and get personalized AI skill recommendations.
Take free scan