// Autonomous GitHub community learning engine — monitor, engage, track, cultivate, report.
| name | github_community |
| description | Autonomous GitHub community learning engine — monitor, engage, track, cultivate, report. |
| tier | lazy |
| input_type | command |
| output_type | report |
| platform | all |
Autonomous flywheel that learns from top GitHub AI communities, engages with genuine expertise, compounds knowledge into DDD, and produces weekly HTML reports. Learning-first — visibility is the byproduct.
TRIGGER: "github community", "community engine", "check github", "community report", "github engage", "github 社区", "看看 github", "community scan". DO NOT USE: for GitHub repo research (use github-research), for GitHub trending (use github-trending), for creating PRs or issues on our own repos (use workspace-git). SIBLINGS: github-research = deep single-repo analysis | github-trending = daily trending snapshot | github-community = ongoing community engagement flywheel.
Read INSTRUCTIONS.md for full workflow.
Orchestrate the full AIDLC Autonomous Pipeline from a one-sentence requirement to a PR-ready delivery with TDD methodology. DDD drives judgment (should we?), SDD produces specs (what exactly?), TDD verifies delivery (did we?). Drives stages sequentially: Evaluate, Think, Plan, Build (TDD red-green), Review, Test, Deliver (with report), Reflect. Classifies every decision as mechanical (auto-approve), taste (batch at delivery gate), or judgment (block for human). Checkpoints on L2 BLOCK, retry exhaustion, or context budget limits. Resumes from checkpoint in a fresh session. TRIGGER: "run pipeline", "autonomous pipeline", "pipeline for", "full pipeline", "build end-to-end", "execute pipeline", "resume pipeline", "continue pipeline", "pipeline status". DO NOT USE: for a single stage (use the specific skill: evaluate, deep-research, code-review, qa, deliver). Not for tasks without a clear requirement.
Unified Account 360 analysis for CMHK customers. Combines company profile, AWS revenue, pipeline, and call plan into one 4-tab HTML report. Uses Sentral MCP for real-time account/opportunity data, DataProxy SDK for Athena revenue, PhoneTool for people, and web search for news. TRIGGER: "account 360", "客户全景", "分析 XX 客户", "XX 客户分析", "call plan for XX", "account analysis". DO NOT USE: for weekly revenue reports (use cmhk-weekly-report), for GTM industry analysis (use cmhk-industry-gtm-analysis), for pipeline-only queries (use cmhk-data-proxy).
BMS Workload Risk & Alert intelligence for CMHK leadership. Queries fd_workload_risk_integrated and app_risk_indicators_*_alert from Athena. Outputs: Risk summary by BU, competitor threat analysis, alert heatmap, account-level risk detail. Self-contained — does NOT depend on other CMHK skills. TRIGGER: "BMS risk", "workload risk", "risk report", "alert heatmap", "at risk accounts", "风险报告", "workload 风险", "BMS 数据". DO NOT USE: for revenue reports (use cmhk-weekly-report), forecast (use cmhk-forecast-report), or account-360 (use cmhk-account-360).
Research customer AI Agent scenarios via web search. Outputs structured JSON/Markdown with each customer's AI products, Agent capabilities, and competitive positioning. TRIGGER: "research AI scenarios", "客户AI调研", "research [company] AI capabilities". DO NOT USE: for revenue data (use s_industry-gtm-analysis) or weekly reports (use s_cmhk-weekly-report).
Generate daily sales briefing reports for CMHK leadership. 5 formats (A-E) with role-based auto-selection, anomaly detection, and AI-generated insights. CSS-only HTML templates. TRIGGER: "daily brief", "日报", "daily push", "sales briefing", "daily report for CMHK". DO NOT USE: for weekly reports (use cmhk-weekly-report), for account deep-dives (use cmhk-account-360), for forecast analysis (use cmhk-forecast-report when available).
Query CMHK revenue, usage, forecast, and account data via the DataProxy API. Covers Athena (fact_estimated_revenue), GCR Sales Data API (forecast cycles, baselines, inputs), hierarchy permissions, and report publishing. TRIGGER: "revenue", "usage", "CMHK data", "weekly numbers", "forecast", "BU revenue", "GenAI revenue", "top accounts", "RFHC numbers", "周报数据", "收入", "用量". DO NOT USE: for generating full HTML reports (that stays on DataRetriever cron), for non-GCR data, or for Sentral/SFDC opportunity data (use aws-sentral-mcp).