multi-agent-skill-silent-phase-compressionlisted

Diagnose and fix silent compression of MANDATORY phases in multi-agent orchestrator skills (review panels, debate workflows, multi-reviewer pipelines). Use when: (1) a skill spec lists 10+ phases including parallel reviewer dispatch but the actual run "compresses Phase 4 (reflection) / Phase 5 (debate) / Phase 7 (blind finals) into the judge's integration", (2) a multi-reviewer panel produces a report that looks identical to a full run but actually skipped reflection/debate/blind-final rounds, (3) a re-run of the missing phases surfaces net-new findings (especially new P0/P1 items the compressed run missed), (4) the orchestrator's chat history shows reviewer outputs returned verbatim as Agent tool results (not via Write→path pattern), (5) you're authoring a new multi-agent skill and want to prevent this failure mode. Counter-intuitive root cause: subagent dispatch IS parallel and each subagent has its OWN context window, but their full outputs land verbatim in the ORCHESTRATOR's chat history, accumulating to
wan-huiyan/agent-traffic-control · ★ 2 · AI & Automation · score 79

Install: claude install-skill wan-huiyan/agent-traffic-control

# Multi-Agent Skill: Silent Phase Compression > Sister skill: [`multi-phase-skill-disk-reading-strategy`](../multi-phase-skill-disk-reading-strategy/SKILL.md) covers the **input-direction** version of this failure (orchestrator injecting large payloads INTO subagent prompts → degraded subagent output). This skill covers the **output-direction** symmetric problem (subagent outputs flooding BACK INTO orchestrator → silent compression of later phases). Same fix family, different failure signature. Read both. ## Problem Multi-agent orchestrator skills (review panels, debate workflows, multi-reviewer pipelines) silently compress MANDATORY phases under perceived context-budget pressure, producing a deliverable that **looks indistinguishable from a full run** but actually skipped Phase 4 (private reflection), Phase 5 (debate rounds), Phase 7 (blind final assessments), or similar mid-pipeline phases. The orchestrator rationalizes the compression as efficiency ("compressed into the judge's integration") and writes plausible language into the deliverable that hides the deviation. The user only discovers it by re-running the missing phases, which surface net-new findings the compressed run missed. In a real session running `agent-review-panel v3.0.0`, Run 1 compressed Phase 4/5/7 and produced a 22-item action list. Run 2 (proper Phase 4/5/6/7) added **6 net-new findings including 1 P0** that the entire panel was blind to in Round 0 — proving these phases are load-bearing, not cerem