{"id":51290101,"url":"https://github.com/nyechiel/ai-augmented-workflow","last_synced_at":"2026-06-30T09:30:44.352Z","repository":{"id":364930521,"uuid":"1269352380","full_name":"nyechiel/ai-augmented-workflow","owner":"nyechiel","description":"A template for building your own AI-powered knowledge worker workflow using Claude Code and MCP","archived":false,"fork":false,"pushed_at":"2026-06-22T06:22:44.000Z","size":143,"stargazers_count":0,"open_issues_count":0,"forks_count":1,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-06-22T08:14:15.015Z","etag":null,"topics":["agent-skills","ai","claude-code","knowledge-worker","mcp","model-context-protocol","productivity","workflow-automation"],"latest_commit_sha":null,"homepage":"https://nyechiel.com/blog/2026/06/14/build-your-own-ai-augmented-workflow/","language":"HTML","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/nyechiel.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":"SECURITY.md","support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-06-14T15:55:53.000Z","updated_at":"2026-06-22T06:22:48.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/nyechiel/ai-augmented-workflow","commit_stats":null,"previous_names":["nyechiel/ai-augmented-workflow"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/nyechiel/ai-augmented-workflow","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nyechiel%2Fai-augmented-workflow","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nyechiel%2Fai-augmented-workflow/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nyechiel%2Fai-augmented-workflow/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nyechiel%2Fai-augmented-workflow/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/nyechiel","download_url":"https://codeload.github.com/nyechiel/ai-augmented-workflow/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nyechiel%2Fai-augmented-workflow/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34961541,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-30T02:00:05.919Z","response_time":92,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agent-skills","ai","claude-code","knowledge-worker","mcp","model-context-protocol","productivity","workflow-automation"],"created_at":"2026-06-30T09:30:43.415Z","updated_at":"2026-06-30T09:30:44.336Z","avatar_url":"https://github.com/nyechiel.png","language":"HTML","funding_links":[],"categories":[],"sub_categories":[],"readme":"# AI-Augmented Workflow\n\n**Give your AI agent full context about your work - email, calendar, chat, tasks, documents - and let it execute multi-step workflows on your behalf, with you as the approver.**\n\n[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)\n[![Platform: Linux | macOS | WSL2](https://img.shields.io/badge/Platform-Linux%20%7C%20macOS%20%7C%20WSL2-lightgrey.svg)](#platform-notes)\n[![Built with Claude Code](https://img.shields.io/badge/Built%20with-Claude%20Code-blueviolet.svg)](https://www.anthropic.com/claude-code)\n[![MCP](https://img.shields.io/badge/MCP-Model%20Context%20Protocol-orange.svg)](https://modelcontextprotocol.io/)\n[![skillsaw grade](https://img.shields.io/endpoint?url=https%3A%2F%2Fraw.githubusercontent.com%2Fnyechiel%2Fai-augmented-workflow%2Fmain%2F.skillsaw-badge.json)](https://skillsaw.org/)\n[![Agent Skills](https://img.shields.io/badge/Agent%20Skills-Standard-green.svg)](https://agentskills.io/home)\n\nA template you clone, customize, and make your own. Built on [Claude Code](https://www.anthropic.com/claude-code) and [MCP](https://modelcontextprotocol.io/), but the architecture - skills as markdown, memory files, tool allowlists - is portable to any MCP-compatible agent.\n\n## Table of contents\n\n- [What's included](#whats-included)\n- [Architecture](#architecture)\n- [Quick start](#quick-start)\n- [Included skills](#included-skills)\n- [Key design decisions](#key-design-decisions)\n- [Important considerations](#warning-important-considerations)\n- [Adapting to your role](#adapting-to-your-role)\n- [Platform notes](#platform-notes)\n- [Further reading](#further-reading)\n- [Contributing](#contributing)\n- [Acknowledgments](#acknowledgments)\n- [License](#license)\n\n## What's included\n\n| Layer | What it does | Key files |\n|-------|-------------|-----------|\n| **Rules** | Defines agent behavior, boundaries, and knowledge | `CLAUDE.md.example` |\n| **Infrastructure** | Connects Claude Code to your tools via MCP | `local-mcp-stack/`, `.mcp.json.example` |\n| **Skills** | Reusable multi-step workflows (inbox triage, meeting prep, etc.) | `skills/` |\n| **Memory** | Persistent cross-session context about you and your work | `memory/` |\n| **Templates** | Starting points for recurring documents | `templates/` |\n| **Ops** | Task board, backup, recovery, and service management | `crux/`, `scripts/`, `app-dashboard/` |\n\n## Architecture\n\n```\nYou (human-as-approver)\n  |\n  v\nClaude Code (agent)\n  |\n  +-- CLAUDE.md (rules, boundaries, knowledge)\n  +-- Skills (inbox-triage, meeting-prep, capture, delegate, review, learn)\n  +-- Memory (user profile, feedback, project context, references)\n  |\n  +-- App Dashboard (localhost:9000 - manages all services below)\n  |\n  +-- MCP Proxy Gateway (localhost:9090)\n  |     +-- Google Workspace (Gmail, Calendar, Drive, Docs)\n  |     +-- Slack (messages, threads, unreads, search)\n  |     +-- Google Contacts (directory lookup)\n  |     +-- Time (timezone-aware clock)\n  |     +-- ... (add your own)\n  |\n  +-- Crux (task board - stdio MCP)\n```\n\nThe proxy gateway aggregates all MCP servers behind a single endpoint with **tool allowlists** - your safety net for enforcing read-only access even when servers have write capabilities.\n\n## Quick start\n\n```bash\ngit clone https://github.com/YOUR_USERNAME/ai-augmented-workflow.git\ncd ai-augmented-workflow\n./scripts/setup.sh        # Linux/macOS/WSL2\n# or: .\\scripts\\setup.ps1  # Windows PowerShell\n```\n\nThe setup script installs Crux and App Dashboard, copies example configs, creates symlinks, and tells you what's left to do manually (MCP server setup, credentials, customization).\n\nSee [docs/SETUP.md](docs/SETUP.md) for the full step-by-step guide.\n\n## Included skills\n\nThese are starter skills to get you up and running. They cover common knowledge worker workflows, but the real value comes from building your own for whatever you repeat. A skill is just a markdown file with instructions - if you can explain it to a person, you can encode it as a skill.\n\n| Skill | What it does |\n|-------|-------------|\n| `/inbox-triage` | Triage unread Slack and Gmail, categorize into respond/review/FYI/defer, extract action items |\n| `/meeting-prep` | Generate a pre-meeting brief with context, prior interactions, and talking points |\n| `/capture` | Auto-discover meeting notes, capture facts/links/contacts, extract action items |\n| `/delegate` | Pick up a queued task, execute it, hand back for review |\n| `/review` | End-of-day or end-of-week review with work output, wins, and risks |\n| `/learn` | Analyze corrections from the session and codify them into memory |\n\nEach skill is a SKILL.md file with a tool allowlist and workflow steps. See [skills/README.md](skills/README.md) for how to create your own.\n\n## Key design decisions\n\n**Human-as-approver.** The agent drafts, researches, and prepares - you decide and send. No messages are ever sent automatically. No issues created without confirmation. This isn't a limitation; it's the trust model that makes the system usable for real work.\n\n**Read-mostly access.** MCP servers are configured with tool allowlists that restrict the agent to read-only access for most integrations. Write access is limited to task management (Crux) and email label management (archive/organize). The proxy config (`mcp-proxy-config.json`) is where you control this.\n\n**Local-first.** Everything runs on your machine. No cloud services beyond the APIs you're already using (Gmail, Slack). Your knowledge base, task history, and memory stay in a git repo you control.\n\n**Skills over prompts.** Instead of re-explaining workflows every session, skills encode them as reusable SKILL.md files with explicit tool allowlists and step-by-step instructions. The agent follows the skill; you review the output.\n\n**Harness over code.** The rules, tool allowlists, memory, and skills form a harness around the agent - guides that steer it before it acts, sensors that correct it after. The harness is the durable artifact; the agent's output is regenerable.\n\n## :warning: Important considerations\n\nThis system connects an AI agent to your real email, calendar, chat, and task management. While it's designed with safety boundaries (tool allowlists, read-only defaults, human-as-approver), you are responsible for understanding what access you're granting and reviewing what the agent does with it.\n\nA few things to keep in mind:\n\n- **Credentials are sensitive.** OAuth tokens and API keys in `mcp-secrets.env` grant access to your accounts. Treat this file like a password. Never commit it to git.\n- **Review tool allowlists carefully.** The proxy config (`mcp-proxy-config.json`) controls which MCP tools are exposed. Before starting the stack, verify you're comfortable with what's enabled - especially any write operations.\n- **Enterprise environments have rules.** If you're connecting to corporate systems (work email, company Slack, internal trackers), check with your IT and security teams first. Many organizations have policies about third-party tool access, API token usage, data handling, and AI tool adoption that may apply.\n- **AI output requires judgment.** The agent can misread context, misattribute information, or produce plausible-sounding errors. The human-as-approver model exists because human review is not optional - it's the safety mechanism.\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ch2\u003eAdapting to your role\u003c/h2\u003e\u003c/summary\u003e\n\nThis template works for any knowledge worker role. What changes:\n\n- **Your MCP servers** - Connect the tools you actually use (Jira, Salesforce, HubSpot, Notion, Confluence, GitHub, etc.)\n- **Your skills** - Build workflows for what you repeat. A few examples by role:\n  - **Sales/account management:** deal prep, pipeline review, account research, RFP drafting\n  - **Marketing:** campaign brief, content calendar review, competitor monitoring\n  - **Engineering:** PR review digest, on-call handoff, incident summary\n  - **Research/academia:** literature review, grant progress tracker, lab notebook capture\n  - **Operations/program management:** status report, risk register update, vendor review\n- **Your memory** - Seed it with your role, domain, and preferences\n- **Your templates** - Documents you create regularly (proposals, reports, briefs, post-mortems, etc.)\n\nSee [docs/CUSTOMIZATION.md](docs/CUSTOMIZATION.md) for a detailed guide, and [workflow-guide.md](workflow-guide.md) for how the skills connect and a suggested daily rhythm.\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ch2\u003ePlatform notes\u003c/h2\u003e\u003c/summary\u003e\n\nThis template is tuned for Linux (Fedora), but the core workflow (Claude Code, MCP servers, skills, memory) works on any platform. The OS-specific parts are the infrastructure layer:\n\n| Component | Linux (included) | macOS | Windows |\n|-----------|-----------------|-------|---------|\n| Containers | Podman + podman-compose | Docker Desktop or Podman Machine | Docker Desktop or WSL2 + Podman |\n| Service management | systemd user services | launchd plist files | Task Scheduler or WSL2 systemd |\n| Backup timer | systemd timer | launchd plist with `StartCalendarInterval` | Task Scheduler |\n| Shell scripts | bash (native) | bash (native) | WSL2 or Git Bash |\n\nIf you're on macOS or Windows, swap the systemd service files for your platform's equivalent and use Docker instead of Podman. Everything else works as-is.\n\n\u003c/details\u003e\n\n## Further reading\n\nThis workflow grew out of daily use over several months. The companion blog post walks through the concepts and what I learned building it:\n\n- [Build Your Own AI-Augmented Workflow](https://nyechiel.com/blog/2026/06/14/build-your-own-ai-augmented-workflow/) - concepts, architecture, and lessons learned\n- [Building an AI Operating System for Product Management](https://nyechiel.com/blog/2026/05/13/ai-operating-system-for-product-management/) - the earlier post that describes the system from the user's perspective\n\n## Contributing\n\nContributions are welcome - bug fixes, documentation improvements, setup and platform compatibility patches. New skills belong in your fork (this is a template, not a shared library). See [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines and [SECURITY.md](SECURITY.md) for reporting vulnerabilities.\n\n## Acknowledgments\n\nThis project builds on ideas and tools from several people:\n\n- [Andrej Karpathy](https://karpathy.ai/) - His [LLM Wiki](https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f) concept (compiling raw data into structured, interlinked markdown rather than chunking it into vector embeddings) is the knowledge architecture pattern behind this workflow. His observations on LLM coding pitfalls, [codified by the community](https://github.com/multica-ai/andrej-karpathy-skills), also helped establish structured project instructions as a best practice.\n- [Anthropic](https://www.anthropic.com/) - Claude Code, the skills system, and the original creators of the [Model Context Protocol](https://modelcontextprotocol.io/), now an open standard under the [Linux Foundation's Agentic AI Foundation](https://www.linuxfoundation.org/press/linux-foundation-announces-the-formation-of-the-agentic-ai-foundation). The skills in this repo follow the [Agent Skills](https://agentskills.io/home) open standard for defining reusable agent capabilities.\n- [TBXark](https://github.com/TBXark) - [mcp-proxy](https://github.com/TBXark/mcp-proxy), the gateway that makes tool allowlists possible - a key safety component.\n- [Boris Cherny](https://github.com/bcherny) - Creator of Claude Code. His [workflow tips](https://howborisusesclaudecode.com/) and approach to CLAUDE.md as a living team document were an early reference for structuring agent instructions.\n- [Jonathan Zarecki](https://github.com/jonzarecki) - His work on a similar AI-augmented workflow was a direct inspiration for this project.\n\n## License\n\n[MIT](LICENSE). Clone it, customize it, build on it - no restrictions beyond preserving the license notice.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnyechiel%2Fai-augmented-workflow","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fnyechiel%2Fai-augmented-workflow","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnyechiel%2Fai-augmented-workflow/lists"}