Expand post-review-pickup trigger surface (Discussion #11423 Phase B Option B.1-prime)

Context Discussion #11423 ideated a mechanism to eliminate the "Helpful Assistant" deference-slip that agents exhibit when transitioning into an idle/blocked state (such as waiting for a peer review or finishing an implementation).

The Problem Agents default to their underlying RLHF conditioning at turn boundaries ("Would you like me to...", "What should I do next?") when a local task finishes, violating the Flat Peer-Team constraint (AGENTS.md §15.6). Descriptive L1 anchor text is too weak against this conditioning.

The Architectural Reality Option B.1-prime was unconditionally [GRADUATION_APPROVED] cross-family. Instead of creating a new state-transition skill (which causes sprawl) or bloating AGENTS.md §0 with universal hard gates, we will expand the post-review-pickup skill's trigger surface to encompass other post-lifecycle-event transitions. This leverages the Phase B Description-Router hardening (PR #11424) to keep the L1 anchor lean.

The Fix

• Extend post-review-pickup-workflow.md description and payload to encompass broader lifecycle transitions.
• Add a highly compact AGENTS.md §21 trigger pointer.
• Update SKILL.md description to explicitly surface the broader trigger language to the description-router.
• Explicitly protect blocked-task-state for strictly negative paths.

Discussion Criteria Mapping Maps directly to the Option B.1-prime [RESOLVED_TO_AC] criteria established in Discussion #11423.

Acceptance Criteria

• Extend .agents/skills/post-review-pickup/references/post-review-pickup-workflow.md description/payload to cover:
  • Post-review handoff (existing)
  • Post-author-response handoff (existing)
  • Post-implementation completion (new)
  • Post-PR-open/update lifecycle event (new)
  • Post-ticket-create lifecycle event (new)
  • Post-blocked-state-resolution (new — reuses blocked-task-state exit signal)
• Add explicit AC rule in the payload: at each broadened lifecycle boundary, agent MUST emit a lane-state: declaration (positive next-lane OR halt with survey evidence).
• Add compact AGENTS.md §21 trigger pointer (1-line max; per ADR 0008 Map-vs-Atlas).
• Update .agents/skills/post-review-pickup/SKILL.md description to surface the broader trigger language.
• Preserve blocked-task-state scope for negative paths only.

Out of Scope

• Creating a new generic state-transition skill.
• Adding universal AGENTS.md hard gates for state transitions.

Related

• Graduates Discussion #11423
• Leverages PR #11424

Origin Session ID 188acb85-b41e-435c-94ee-0cc9944d4c97