@seanford

web-retrieval wraps scrapling for static html fetch, js-rendered pages, and anti-bot sites via three escalating fetchers (curl_cffi, dynamic, stealthy). supports bulk url crawls with checkpoint/resume and css selector extraction.

skilled-openclaw-advisor queries a local sqlite fts5 index of openclaw documentation for configuration, cli commands, and troubleshooting. no external api calls, sub-10ms latency. three query modes (agent, human standard, human detailed) and optional language/verbosity config.

nerve-kanban exposes rest endpoints for task lifecycle management on a kanban board, including crud operations, workflow transitions (execute, approve, reject, abort), proposal handling, and board configuration. version-based conflict resolution required for mutations.

skilled-deep-research coordinates parallel sub-agents across gov/github/reddit/news domains to fetch and synthesize cited sources. Uses acp spawning, mcp browser tools, and a2a signaling for checkpoint-driven research runs at three tiers (simple/standard/deep).

ada-agent-autonomy-kit enables agents to execute queued tasks without explicit prompts via heartbeat polling and cron scheduling, shifting from reactive to continuous operation.

Consult the tool-models agent for model selection, cost estimation, capability queries, and provider comparisons. Use when choosing a model for any task, com...