@sunnyhot

autonomous-pm coordinates decentralized multi-agent work via shared STATE.yaml files, eliminating central orchestration bottlenecks. agents track task status, dependencies, and execution history through git-versioned state, enabling parallel execution without polling.

podcast-production-pipeline automates content workflow from topic selection through multi-platform publishing, with domestic chinese platform adapters. lacks explicit error handling and recovery procedures.

hot-topics-daily fetches trending topics from six chinese platforms (weibo, zhihu, baidu, bilibili, douyin, toutiao) and posts them daily to discord. supports per-platform toggles and configurable item counts.

failure-monitor detects cron job failures, diagnoses error types (timeout, permission, api key, network), and auto-repairs simple config issues while escalating permission/script changes to human review via discord notifications.

skill-sync monitors local skill directories for changes and automatically publishes updates to clawhub and github. detects new or modified skills and maintains sync state, but lacks explicit failure handling and recovery procedures.

openclaw-log-analyzer monitors cron job and system logs to detect errors, analyze patterns, and push discord reports on daily/weekly cycles. lacks clear decision logic and failure recovery paths.

sunny-health-monitor collects cpu, memory, disk, and cron job metrics on macos, calculates a health score, and posts formatted reports to discord every 30 minutes. lacks actionable procedure detail and error handling.

opik-reporter fetches traces from opik observability platform, computes token usage and costs, identifies errors, then sends formatted summaries to discord. documentation is in chinese and lacks actionable procedural depth.

idea-reality-validator scans github, hacker news, npm, pypi, and product hunt to assess idea competition before development. outputs a 0-100 reality signal with top competitors and differentiation suggestions, including china market sources.

daily-self-improvement collects error logs and daily notes, analyzes problem patterns, searches clawhub for relevant skills, and generates a markdown report pushed to discord nightly. execution is vague on actual implementation details.

agent-swarm-dev-team coordinates multiple coding agents (codex, claude, gemini) across isolated git worktrees with tmux session management and cron-based monitoring, though critical implementation details remain unspecified.

deals-hunter aggregates what's worth buying rss feeds with tavily-powered price history lookups (capped at 5 items per run) and purchase recommendations. includes 24-hour caching and three daily scheduled triggers.

semantic-memory-search adds vector-driven search to openclaw markdown memory files using the memsearch library with hybrid search (dense vectors plus bm25) and local embedding without api keys.

polymarket-autopilot claims to automate paper trading via ai agents with market monitoring and portfolio management, but lacks concrete procedural steps, decision logic, and failure handling. implementation details are absent.

A comprehensive ClawHub skill installer that bypasses API rate limits. Search, install single skills, or batch install multiple skills with automatic retry l...

A comprehensive ClawHub skill installer that bypasses API rate limits. Search, install single skills, or batch install multiple skills with automatic retry l...

Automatically monitor and claim ClawPI red packets. Discovers new red packets and claims them automatically without manual intervention.

AI 驱动的财报追踪器，自动扫描 A 股/美股财报日历，推送重要财报更新

Professional health assistant with medical report analysis, evidence-based supplement recommendations (100+ supplements), machine learning personalization, e...