440 OneWave-AI
agent-swarm-deployer Skill
部署子代理集群,处理大规模并行数据处理任务。与面向代码修改的 agent-army 不同,它专用于数据任务——处理 1000 份文档、分析数据集、批量内容生成。支持可配置的集群规模、任务分发、结果聚合、进度跟踪和错误恢复。
安装方式:把技能目录放入 ~/.claude/skills/(Claude Code)或在 claude.ai 设置中启用;也可复制右侧安装命令一键添加。
技能指令原文(SKILL.md)
Agent Swarm Deployer
Deploy a swarm of parallel sub-agents to process massive, independent data tasks (documents, records, rows, items) and aggregate the results. Use this for data operations; use agent-army for code changes.
Contents
references/overview.md-- swarm vs army comparison, use cases, architecture diagramreferences/swarm-design.md-- input/output schemas, batch-size and swarm-size formulas, scaling guidelinesreferences/agent-brief.md-- agent brief template, data distribution methods, progress trackingreferences/aggregation-recovery.md-- merge logic, completeness validation, retry strategy, error-handling tablereferences/output-formats.md-- CSV/JSON/Markdown/individual-file outputs, final summary reportreferences/task-configs.md-- ready-made configs for sentiment, lead scoring, content generation, summarization
Workflow
- Understand the task. Pin down five things before deploying anything: data source, operation per item, output format, output destination, and quality/validation requirements. If any is ambiguous, ask the user first -- a wrong spec wastes all agent compute.
- Intake and inventory. Glob/Bash to locate and count items. Read 3-5 samples to learn structure. Estimate tokens per item and total. Report an intake summary (source, total count, item format, sample structure, token estimate).
- Detect input schema and define output schema. Derive the input schema from samples; define the exact output schema the task requires. See
references/swarm-design.md.
- Design the swarm. Compute batch size from token budget (70% of ~200K usable context per agent) and swarm size from total items. Cap at 20 agents per wave; split into waves beyond that. Present the swarm plan and agent assignments, then get approval. See
references/swarm-design.md.
- Prepare agent briefs. Build a self-contained brief per agent: role, task, input data, output schema with example, quality rules, error protocol, and strict JSON output format. See
references/agent-brief.md.
- Distribute data and deploy. Choose a distribution method for the source type (pre-split CSVs/JSON with Bash; embed inline for small sets; pass file paths for directories). Launch up to 20 agents in parallel via the Agent tool with
run_in_background: true, sending all calls in one message. Run later waves after the prior wave completes. Seereferences/agent-brief.md.
- Track progress. As agents return, record status, processed counts, and cumulative coverage. See
references/agent-brief.md.
- Collect and aggregate. Parse each agent's JSON; validate schema, completeness, and duplicates. Merge into one ordered output and extract failures. Report an aggregation summary with a coverage check and failure analysis. See
references/aggregation-recovery.md.
- Recover failures. Queue all failed and skipped items, deploy a retry agent with enhanced instructions, cap at 2 retries, and mark survivors "unrecoverable". Flag the user if unrecoverable items exceed 10%. See
references/aggregation-recovery.md.
- Write output and summarize. Produce the requested format (CSV, JSON, Markdown, or individual files) plus a final summary covering execution, results, quality metrics, patterns observed, and cost. See
references/output-formats.md.
Anti-Patterns to Avoid
- Do not use a swarm for sequential tasks. If item N depends on item N-1, use a chain instead.
- Do not deploy one agent per item. Batch items; one-per-agent wastes overhead.
- Do not skip schema definition. Without a schema, merging results from many agents becomes unreliable.
- Do not ignore failures. At 99% success, 1% of 10,000 items is still 100 failures. Always run retries.
- Do not deploy without a sample run. Process 5 items manually first to validate the task and output quality before scaling.