aiwg
Version:
Deployment tool and support utility for AI context. Copies agents, skills, commands, rules, and behaviors into the paths each AI platform reads (Claude Code, Codex, Copilot, Cursor, Warp, OpenClaw, and 6 more) so one source of truth works across 10 platfo
44 lines • 1.74 kB
TypeScript
/**
* RLM Agentic Tools CLI — Subcommand router for RLM support tools
*
* Subcommands:
* chunk <file> — Split file into overlapping chunks for fanout
* fanout <query> — Dispatch parallel subagent queries across chunks
* rlm-prep <file|dir> — Prepare source content (chunk + index + manifest)
* rlm-search <query> — Full recursive search pipeline
* rlm-status — Show active task tree, progress, and cost
*
* These tools are designed for agentic sessions implementing the RLM pattern
* (recursive decomposition + programmatic environment interaction). They are
* also directly invocable by users.
*
* Research foundation: REF-089 (Zhang et al., 2026)
*
* @implements #559
*/
/**
* Resolve the effective `--parallel` / `--max-parallel` value for RLM CLI
* commands, composing all applicable caps. Returns:
* { effective, source, clamped, hardCapHit }
*
* Precedence (smallest wins):
* 1. `parallelism.max_parallel_subagents` from `.aiwg/aiwg.config` (#1359)
* 2. The RLM Rule 8 hard cap of 7
* 3. The user-supplied flag value (or hardcoded fallback when unset)
*
* When the user passes a value above the cap, we warn and clamp. When no
* flag is passed, we use the resolved cap directly as the default — that is
* the whole point of the project-level config.
*
* @implements #1360
*/
export declare function resolveRlmParallel(userValue: number | undefined, fallbackDefault: number, projectDir?: string): Promise<{
effective: number;
source: string;
warning?: string;
}>;
/**
* Main CLI entry point for RLM agentic tool subcommands
*/
export declare function main(args: string[]): Promise<void>;
//# sourceMappingURL=cli.d.ts.map