By ianliuy
Lightweight Teams message skill: draft with context → fail-closed AI review (Copilot CLI / Codex) → send to self via Teams MCP. v1 is self-send only.
Lightweight plugin for sending Teams messages with AI-assisted drafting and fail-closed AI review.
Works with both Copilot CLI and Claude Code (same plugin structure).
v1 scope: self-send only — messages are sent to your own "Notes to Self" chat via Teams MCP. You review and manually forward to the intended recipient.
User context → AI draft (recipient + HTML message + reason) → AI review (fail-closed) → Send to self via Teams MCP
# Invoke the skill:
/genflow-teams:send remind Alice that tomorrow's standup is moved to 10am
# With more context:
/genflow-teams:send tell Alice the PR feedback has been addressed, ask for re-review
Add to ~/.copilot/mcp-config.json (Copilot CLI) or .vscode/mcp.json (VS Code):
{
"mcpServers": {
"microsoft-teams": {
"url": "https://agent365.svc.cloud.microsoft/agents/tenants/72f988bf-86f1-41af-91ab-2d7cd011db47/servers/mcp_TeamsServer",
"type": "http"
}
}
}
Auth uses your Microsoft tenant identity automatically. The tenant ID 72f988bf-... is Microsoft's public tenant — this is the correct value for Microsoft corp accounts.
See Microsoft MCP documentation for setup details.
The review gate uses Copilot CLI (copilot -p -s) if available, falls back to Codex CLI (codex exec). If neither is found, the skill will NOT send the message.
npm install -g @openai/codex├── .claude-plugin/plugin.json # Plugin metadata
├── commands/send.md # /genflow-teams:send command entry
├── skills/teams-send/SKILL.md # Core skill (flow: draft → review → send)
├── scripts/
│ ├── review-message.sh # AI review gate (Copilot CLI → Codex fallback)
│ └── review-message.cmd # Windows launcher
├── config/
│ └── default_config.json # Defaults (tone, codex model)
└── .genflow/teams/
├── drafts/ # Message drafts (md files)
├── history/ # Audit trail (JSON per invocation)
└── reviews/ # Review results
The review gate is fail-closed (uses Copilot CLI, falls back to Codex):
APPROVED → send to selfREVISION_NEEDED: <suggestions> → revise and retry (max 2 retries)REJECT: <reason> → stop, do not sendThere is no auto-approve. If both Copilot CLI and Codex are down, the message is not sent.
Every invocation (sent, rejected, or failed) produces a JSON file in .genflow/teams/history/ with: run_id, timestamp, intended_recipient, message_html, reason, review_result, status, teams_message_id.
teams-send, the single capability within this pluginOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub ianliuy/skill-gen-flow-teamsHumanize - An iterative development plugin that uses Codex to review Claude's work. Creates a feedback loop where Claude implements plans and Codex independently reviews progress, ensuring quality through continuous refinement.
Ultra-compressed communication mode. Cuts 65% of output tokens (measured) while keeping full technical accuracy by speaking like a caveman.
Comprehensive UI/UX design plugin for mobile (iOS, Android, React Native) and web applications with design systems, accessibility, and modern patterns
Standalone image generation plugin using Nano Banana MCP server. Generates and edits images, icons, diagrams, patterns, and visual assets via Gemini image models. No Gemini CLI dependency required.
Multi-model consensus engine integrating OpenAI Codex CLI, Gemini CLI, and Claude CLI for collaborative code review and problem-solving.
Unified capability management center for Skills, Agents, and Commands.
Write feature specs, plan roadmaps, and synthesize user research faster. Keep stakeholders updated and stay ahead of the competitive landscape.