Decompose large inputs into chunks processed by parallel Haiku sub-agents and synthesized by a Sonnet supervisor.
npx skills add https://github.com/mifunedev/orchestra --skill rlm-mifunedev
Please help me install this Agent Skill. Skill: rlm Source: https://github.com/mifunedev/orchestra/tree/main/.claude/skills/rlm Command: npx skills add https://github.com/mifunedev/orchestra --skill rlm-mifunedev
Large inputs often exceed single-pass processing limits, requiring a scalable approach to break down tasks, run parallel analyses, and synthesize results.
Invoke the RLM workflow on a large input to decompose it, run parallel Haiku workers, and synthesize the final results.
High-intent search queries and answers about installing and using this skill.
To process large codebases, you can decompose the input into manageable chunks, delegate concurrent processing to parallel Haiku sub-agents, and synthesize the unified results using a Sonnet supervisor.
Analyzing lengthy documents with parallel processing involves breaking the text into chunks, running concurrent Haiku workers on each segment, and having a Sonnet supervisor synthesize the outputs for a unified result.
Yes, you can use Haiku and Sonnet models together for multi-file analyses by delegating chunked execution to parallel Haiku sub-agents while a Sonnet model acts as the supervisor to synthesize the final output.
You can reduce costs in data-heavy workflows by decomposing large inputs into smaller chunks and processing them concurrently with Haiku workers, which speeds up outcomes and lowers processing expenses.
Task-based orchestration for chunked document processing is a pattern that manages the decomposition, parallel execution, and final synthesis of large inputs using a scratchpad, Haiku workers, and a Sonnet supervisor.
ruska-aiDecompose large inputs and coordinate parallel sub-agents for synthesis.
TheBushidoCollectiveExport TensorFlow models to SavedModel and convert to TFLite for deployment.
frankxaiCoordinate multiple Starlight agents for parallel execution and result synthesis.
ArvorCoAutomate TDD-driven feature implementation with quality gates and PRD alignment.
