メニュー
Use when executing implementation plans with independent tasks in the current session. Dispatches fresh subagent per task with two-stage review (spec compliance, then code quality).
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Drift-specific ADR workflow for signal, scoring, output, and architecture decisions. Use when creating, updating, or validating ADRs before implementation. Keywords: ADR, decision record, proposed, signal design, scoring change, architecture boundary, output format, Decision trailer.
Drift-specific workflow for creating, extending, and validating ground-truth fixtures for signal precision and recall. Use when adding TP/TN fixtures, confounders, boundary cases, or file_history_overrides in tests/fixtures/ground_truth.py. Keywords: ground truth, fixture, precision recall, TP, TN, confounder, boundary, file_history_overrides, ALL_FIXTURES.
End-to-end Drift workflow for adding or materially changing a signal. Use when designing a signal, implementing BaseSignal logic, adding weights, creating fixtures, updating audits, and validating precision/recall before commit. Keywords: signal development, BaseSignal, register_signal, incremental_scope, signal design, precision recall, mutation benchmark, audit_results, ADR.
Operationaler Workflow für den internen MCP-Server aus scripts/mcp_product_health_server.py. Immer verwenden, wenn Agenten Produktstatus, Release-Gates, Working-Tree-Gates oder Audit-Freshness direkt per MCP abrufen sollen statt manuelle Git-/Script-Aufrufe zu bauen. Keywords: drift_product_health, drift_release_readiness, drift_working_tree_context, drift_audit_freshness, mcp server, pre-push gates, blast report, audit freshness.
Log learnings, errors, and corrections to .learnings/ for continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects the agent, (3) A knowledge gap is identified, (4) A better approach is found. Captures corrections, insights, errors, and feature requests; promotes broadly applicable learnings to project memory files.
Post-completion self-review for coding agents that runs simplify, harden, and micro-documentation passes on non-trivial code changes. Use when: a coding task just completed, the diff touches executable source files with 10+ changed lines or high-impact logic (auth, validation, queries, file paths, network, concurrency). Feeds recurring findings into self-improving-agent via .learnings/LEARNINGS.md.
| name | subagent-driven-development |
| description | Use when executing implementation plans with independent tasks in the current session. Dispatches fresh subagent per task with two-stage review (spec compliance, then code quality). |
| argument-hint | Provide the path to the implementation plan to execute. |
Execute plan by dispatching fresh subagent per task, with two-stage review after each: spec compliance review first, then code quality review.
Core principle: Fresh subagent per task + two-stage review (spec then quality) = high quality, fast iteration
vs. Executing Plans: Same session, fresh context per task, faster iteration with no human-in-loop between tasks.
superpowers:finishing-a-development-branchNever ignore an escalation or force the same model to retry without changes.
./implementer-prompt.md — Dispatch implementer subagent./spec-reviewer-prompt.md — Dispatch spec compliance reviewer subagent./code-quality-reviewer-prompt.md — Dispatch code quality reviewer subagentNever:
Required workflow skills:
superpowers:writing-plans — Creates the plan this skill executessuperpowers:test-driven-development — Subagents follow TDD for each tasksuperpowers:finishing-a-development-branch — Complete development after all tasks