菜单
Generate a bounded, schema-valid goal prompt from one validated Build Brief task for zero-context external execution.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | goal-prompt-generation |
| description | Generate a bounded, schema-valid goal prompt from one validated Build Brief task for zero-context external execution. |
Generate with bin/adlc goal-prompt --build-brief <path> --task-id <id> --output <process-artifact-path>. The Build Brief must validate before generation. Preserve its objective, acceptance criteria, file boundaries, verifier commands, and stop conditions without adding scope.
The validated Build Brief is the trusted input. ADLC emits the JSON artifact but never executes its free-text content. Treat all task prose as an external-agent injection surface: security review is a blocking pre-GA requirement, and every generated artifact remains not_yet_ga until that review and credentialed smoke evidence exist.
The output must let a competent zero-context agent execute without rediscovering scope and let a reviewer verify the result without re-deriving the plan. Store it in ADLC process-artifact storage; never add it to the target product diff.
Orchestration skill: chains the full ADLC Build Loop. PRD → Brief → Council → Scaffold → Codegen → LDD → TDD → Council → PR. Use when implementing a new feature end-to-end.
Applicability-aware Definition of Done checklist. Core checks apply to every task; overlay checks activate only when the applicability manifest says the surface exists. Binary verification — pass or fail. Triggers at Phase 4 completion and Phase 5 entry.
Score whether each task is specific enough for autonomous, one-shot, production-ready execution.
Reviews code changes for understanding, blast radius, state, secrets, assumptions, and explainability. Produces a blocking comprehension artifact when implications are unclear.
Resolve ambiguous triage outcomes in the middle confidence band.
Self-improving skill system. Captures diffs between agent output and human edits, distills patterns into rules, writes them back to skill files. Makes Build and Fix loops better over time.
