Orchestrate reviewer-gated iterative refinement for parallel agent workers.
npx skills add https://github.com/quickcall-dev/skills --skill iterative-fleet
Please help me install this Agent Skill. Skill: iterative-fleet Source: https://github.com/quickcall-dev/skills/tree/main/skills/iterative-fleet Command: npx skills add https://github.com/quickcall-dev/skills --skill iterative-fleet
iterative-fleet helps teams run complex, multi-round work with multiple parallel agent workers while enforcing a quality gate so the work only advances when an explicit reviewer verdict approves it.
depends_on to run workers in topological layers for correct sequencing (e.g., researcher → builder → reviewer) across every iteration.claude -p, codex exec, or pi -p providers, with configurable model settings per fleet/worker.Launch an iterative reviewer-gated fleet by creating a fleet.json with exactly one type: "reviewer" worker and then running bash ${AGENTS_SKILLS_DIR}/scripts/launch.sh $FLEET_ROOT followed by bash ${AGENTS_SKILLS_DIR}/scripts/status.sh $FLEET_ROOT.
High-intent search queries and answers about installing and using this skill.
You orchestrate reviewer-gated parallel agent workers by defining a fleet.json with one reviewer worker, then running launch.sh and status.sh. The orchestrator coordinates execution via DAG layers and applies stop conditions like max_iterations or reviewer_lgtm_count.
DAG scheduling executes parallel agent workers in topological layers using depends_on to sequence tasks correctly across every iteration. This ensures dependencies like researcher, builder, and reviewer run in the proper order before the next layer starts.
Multiple AI providers are supported for parallel worker execution, including headless claude -p, codex exec, and pi -p. You can configure model settings per fleet or per worker to match your specific task requirements.
The fleet orchestrator never auto-kills or auto-restarts agent workers. It only reads reviewer verdicts and worker completion signals, ensuring operators remain in full control of any manual intervention during iterative refinement.
Stop conditions for iterative agent refinement include max_iterations to cap cycles, reviewer_lgtm_count to stop after enough approvals, and an optional cost_cap_usd limit to control spending across parallel workers.
Tmux automation is used to manage the parallel worker processes within the iterative fleet orchestration. You need a fleet.json configuration defining your workers and a reviewer to initialize the environment before launching the workflow.
quickcall-devOrchestrate AI worker tasks in a DAG with budget caps and tmux panes.
rongaredeParse Codex CLI NDJSON logs to extract execution summaries and failure details.
kwt00Manage parallel CLI worker tasks across tmux panes.
starrank-softOrchestrates autonomous development workflows using sequential or DAG-based agent loops.