Review Studio 2.0

Geo Content Brief Skill

将 GEO 内容访谈、关键词笔记、竞品线索和渠道约束整理成可执行的中文内容简报。用于团队复用、选题规划、内容生产交接和质量评审；不用于直接代写完整长文或替代品牌策略判断。

审查结论 review Score 67/100

核心指标

Skill IR2.0.0

5 targets in platform-neutral contract

Compiler5/5

target contracts compiled from Skill IR

Output Deltan/a

0 cases; 0 file-backed

Exec Runs0

command 0; model 0; recorded 0

Blind A/B0

review pairs hide baseline vs with-skill labels

Review Kit0/0

pending 0; answer key hidden

Review A/B0/0

adjudication decisions; pending 0

Public Claimblocked

0 blockers; local reproducible false

Blueprint0/0

2.0 coverage; evidence pending 0

Runtime0/0

target conformance pass rate

Perm Probe0/0

0 native; 0 installer-enforced

Trust0

0 scripts scanned; secrets found

Py Compat0

0 files scanned for Python 3.11

Arch Debt0

0 largest lines; 0 CLI handlers

Atlas0

0 scanned skills; route collisions

Driftno-data

0 metadata events; 0 missed triggers

Waivers0

0 gates covered; human risk decisions

Intake0/0

0 valid submissions; 0 invalid

Claim Guard0

0 public surfaces scanned

Notes0/0

0 open blocker annotations

Registryn/a

0 targets; no license license

Archiven/a

0 zip entries; package verification

Installn/a

0 adapters; 0 permissions enforced; 0 permission failures

Upgraden/a

declared n/a; 0 breaking changes

审查闸门

关注

意图画布

intent confidence 10/100; Pause before deep authoring and close the highest-leverage gaps first.

reports/intent-confidence.json 证据
关注

触发实验

route scorecard is missing; run route-scorecard before release review

reports/route_scorecard.json 证据
关注

输出实验

output eval scorecard is missing; generate it before production review

reports/output_quality_scorecard.json 证据
关注

上下文

initial load n/a/n/a; deferred n/a/n/a; top deferred none; resource governance unknown; quality density n/a

reports/context_budget.json 证据
关注

运行矩阵

runtime conformance matrix is missing

reports/conformance_matrix.json 证据
关注

信任报告

security trust report is missing

reports/security_trust_report.json 证据
关注

Python 兼容

Python compatibility report is missing

reports/python_compatibility.json 证据
通过

架构维护

architecture maintainability report is optional until code surface grows

reports/architecture_maintainability.json 证据
关注

权限批准

permission governance evidence is missing because trust report is missing

reports/security_trust_report.json + security/permission_policy.json 证据
关注

权限探针

runtime permission probe report is missing

reports/runtime_permission_probes.json 证据
关注

组合治理

skill atlas is missing; portfolio-level conflicts are unknown

reports/skill_atlas.json 证据
关注

运营回路

0 metadata events; adoption 0; missed 0; bad-output 0; risk no-data

reports/adoption_drift_report.json 证据
关注

人工批准

0 active waivers; 11 warning gates still need reviewer decision

reports/review_waivers.json 证据
通过

世界证据

world-class evidence ledger is optional until governed or public world-class readiness is claimed

reports/world_class_evidence_ledger.json 证据
通过

注册审计

registry audit is optional until team distribution is required

reports/registry_audit.json + reports/install_simulation.json 证据
关注

发布路线

promotion decisions are missing; release notes need reviewer confirmation

reports/promotion_decisions.json + reports/upgrade_check.json + docs/migration-v2.md 证据

阻断事项

无。

关注事项

  • 意图画布intent confidence 10/100; Pause before deep authoring and close the highest-leverage gaps first.
  • 触发实验route scorecard is missing; run route-scorecard before release review
  • 输出实验output eval scorecard is missing; generate it before production review
  • 上下文initial load n/a/n/a; deferred n/a/n/a; top deferred none; resource governance unknown; quality density n/a
  • 运行矩阵runtime conformance matrix is missing
  • 信任报告security trust report is missing
  • Python 兼容Python compatibility report is missing
  • 权限批准permission governance evidence is missing because trust report is missing
  • 权限探针runtime permission probe report is missing
  • 组合治理skill atlas is missing; portfolio-level conflicts are unknown
  • 运营回路0 metadata events; adoption 0; missed 0; bad-output 0; risk no-data
  • 人工批准0 active waivers; 11 warning gates still need reviewer decision
  • 发布路线promotion decisions are missing; release notes need reviewer confirmation

修复动作

关注

意图画布

收紧真实任务、输入、输出、排除项和成功标准。

低 intent confidence 会让后续 Skill IR、输出评测和 Review Studio 结论建立在模糊意图上。
修复位置
reports/intent-dialogue.md + reports/intent-confidence.md
验证命令
python3 scripts/yao.py intent-confidence .
关注

触发实验

修正 route scorecard 中的误触发、漏触发或 ambiguous case。

触发错误会让正确 Skill 失活，或让相邻 Skill 被错误调用。
修复位置
SKILL.md frontmatter description + evals/*/trigger_cases.json
验证命令
python3 scripts/build_confusion_matrix.py
  • SKILL.md:3frontmatter description · source
  • evals/trigger_cases.json · missingtrigger eval cases · eval
  • reports/route_scorecard.md · missingroute scorecard · report
关注

输出实验

补足 output eval 覆盖、execution evidence、blind A/B 和 reviewer adjudication。

没有输出质量和人工盲评证据时，Skill 只能证明会触发，不能证明输出真的更好且经得起审查。
修复位置
evals/output/cases.jsonl + reports/output_quality_scorecard.md + reports/output_review_kit.html + reports/output_review_adjudication.md
验证命令
python3 scripts/adjudicate_output_review.py --write-template && python3 scripts/yao.py output-review
  • evals/output/cases.jsonl · missingoutput eval cases · eval
  • reports/output_quality_scorecard.md · missingoutput scorecard · report
  • reports/output_execution_runs.md · missingoutput execution runs · report
  • reports/output_blind_review_pack.md · missingblind A/B review pack · report
  • reports/output_review_kit.html · missingreviewer cockpit · report
  • reports/output_review_adjudication.md · missingreview adjudication · report
关注

上下文

压缩或拆分高成本 deferred resources，保留最小可路由上下文。

初始加载可以安全，但后续 references、scripts、evals 体量过大时，reviewer 仍需要看到维护和读取成本。
修复位置
SKILL.md + references/ + scripts/ + evals/
验证命令
python3 scripts/render_context_reports.py
  • SKILL.md:1entrypoint · source
  • reports/context_budget.md · missingcontext budget · report
  • scripts/resource_boundary_check.py · missingresource boundary checker · source
  • references/skill-engineering-method.md · missingskill engineering method · method
关注

运行矩阵

修复目标端结构、metadata、相对路径、fallback 或 adapter target 声明。

runtime conformance 失败意味着包可能被目标客户端错误加载或静默降级。
修复位置
agents/interface.yaml + reports/conformance_matrix.md
验证命令
python3 scripts/run_conformance_suite.py .
  • agents/interface.yaml:7portable interface · source
  • reports/conformance_matrix.md · missingconformance matrix · report
关注

信任报告

处理脚本 help surface、依赖 pin、network policy、secret 和权限声明。

团队分发时，脚本和依赖是主要供应链风险面，warning 必须有明确处置。
修复位置
reports/security_trust_report.md + security/*.md + scripts/
验证命令
python3 scripts/trust_check.py .
  • reports/security_trust_report.md · missingtrust report · report
  • security/script_policy.md · missingscript policy · policy
  • security/network_policy.md · missingnetwork policy · policy
关注

Python 兼容

修复 Python 3.11 语法兼容问题，尤其是 f-string 表达式内的反斜杠转义。

目标运行环境可能仍停留在 Python 3.11，语法漂移会让 CLI、报告生成和 CI 在发布后直接失败。
修复位置
reports/python_compatibility.md + scripts/*.py + tests/*.py
验证命令
python3 scripts/yao.py python-compat .
  • reports/python_compatibility.md · missingPython compatibility · report
  • scripts/python_compat_check.py · missingcompatibility checker · source
  • .github/workflows/test.yml · missingCI test workflow · ci
关注

权限批准

补齐高权限能力的 reviewer、scope、reason、expires_at 和目标端 enforcement 说明。

权限契约只有在批准人、有效期和目标端处置方式明确时，才能支撑 governed release。
修复位置
security/permission_policy.json + security/permission_policy.md
验证命令
python3 scripts/trust_check.py .
  • security/permission_policy.json · missingpermission approvals · policy
  • security/permission_policy.md · missingpermission method · policy
reports/security_trust_report.json + security/permission_policy.json 打开证据
关注

权限探针

生成并修复目标包的 runtime permission probe 报告。

目标端即使只能提供 metadata fallback，也必须明确 native enforcement 缺口、表示位置和 operator note。
修复位置
dist/targets/*/adapter.json + reports/runtime_permission_probes.md
验证命令
python3 scripts/probe_runtime_permissions.py . --package-dir dist
  • reports/runtime_permission_probes.md · missingruntime permission probes · report
  • dist/targets/openai/adapter.json · missingOpenAI adapter · package
  • dist/targets/claude/adapter.json · missingClaude adapter · package
  • dist/targets/generic/adapter.json · missinggeneric adapter · package
关注

组合治理

处理 portfolio 里的路由冲突、owner 缺口、stale skill 和重复能力。

单个 Skill 质量很高仍可能在团队 skill library 中被相邻 Skill 冲突削弱。
修复位置
reports/skill_atlas.html + skill_atlas/catalog.json
验证命令
python3 scripts/build_skill_atlas.py --workspace-root .
  • skill_atlas/catalog.json · missingskill atlas catalog · atlas
  • skill_atlas/policy.json · missingatlas scope policy · policy
  • reports/skill_atlas.html · missingskill atlas report · report
关注

运营回路

记录 metadata-only 使用事件，或明确当前 release 缺少真实使用信号。

没有运营回路时，reviewer 无法判断采用率、误触发、坏输出和 review overdue 的真实影响。
修复位置
reports/adoption_drift_report.md
验证命令
python3 scripts/render_adoption_drift_report.py . --record-event skill_activation --activation-type explicit --outcome accepted
关注

人工批准

对保留的 warning 写入 reviewer、理由、范围和到期时间，或修掉 warning。

warning 可以被接受，但必须可审计、会过期，并且不能掩盖 blocker。
修复位置
reports/review_waivers.md
验证命令
python3 scripts/render_review_waivers.py .
关注

发布路线

确认 promotion、upgrade diff、breaking changes、migration guide 和 known limitations。

发布说明不完整会让使用者无法判断升级风险和迁移动作。
修复位置
reports/upgrade_check.md + docs/migration-v2.md
验证命令
python3 scripts/upgrade_check.py . --previous-package-json registry/examples/yao-meta-skill-1.0.0.json
  • reports/upgrade_check.md · missingupgrade check · report
  • docs/migration-v2.md · missingmigration guide · docs
  • reports/promotion_decisions.md · missingpromotion decisions · report
reports/promotion_decisions.json + reports/upgrade_check.json + docs/migration-v2.md 打开证据

审查批注

当前批注：0 条批注；开放 0；阻断 0

当前没有 reviewer 批注。

意图画布

将 GEO 内容访谈、关键词笔记和渠道约束整理成可执行的中文内容简报 Skill，适用于团队复用、触发边界审查和后续质量评估。

证据路径

  • skill_overviewreports/skill-overview.html
  • review_viewerreports/review-viewer.html
  • output_evalreports/output_quality_scorecard.md
  • output_executionreports/output_execution_runs.md
  • output_blind_reviewreports/output_blind_review_pack.md
  • output_review_kitreports/output_review_kit.md
  • output_review_kit_htmlreports/output_review_kit.html
  • output_review_decisionsreports/output_review_decisions.json
  • output_review_adjudicationreports/output_review_adjudication.md
  • benchmark_reproducibilityreports/benchmark_reproducibility.md
  • skill_os2_coveragereports/skill_os2_coverage.md
  • runtime_conformancereports/conformance_matrix.md
  • trust_reportreports/security_trust_report.md
  • python_compatibilityreports/python_compatibility.md
  • architecture_maintainabilityreports/architecture_maintainability.md
  • permission_policysecurity/permission_policy.md
  • runtime_permissionsreports/runtime_permission_probes.md
  • skill_atlasreports/skill_atlas.html
  • compiled_targetsreports/compiled_targets.md
  • adoption_driftreports/adoption_drift_report.md
  • review_waiversreports/review_waivers.md
  • review_annotationsreports/review_annotations.md
  • world_class_evidence_planreports/world_class_evidence_plan.md
  • world_class_evidence_ledgerreports/world_class_evidence_ledger.md
  • world_class_evidence_intakereports/world_class_evidence_intake.md
  • world_class_submission_reviewreports/world_class_submission_review.md
  • world_class_operator_runbookreports/world_class_operator_runbook.md
  • world_class_operator_runbook_htmlreports/world_class_operator_runbook.html
  • world_class_claim_guardreports/world_class_claim_guard.md
  • registry_auditreports/registry_audit.md
  • package_verificationreports/package_verification.md
  • install_simulationreports/install_simulation.md
  • upgrade_checkreports/upgrade_check.md
  • migrationdocs/migration-v2.md
  • skill_irreports/skill-ir.json

触发实验

route scorecard is missing; run route-scorecard before release review

组合治理

skill atlas summary missing

输出实验

output eval scorecard missing

执行证据

output execution report missing

盲评包

blind A/B review pack missing

审定报告

review adjudication report missing

评审清单

先打开 reports/output_review_kit.md；每张卡片对应一个 blind A/B case，说明当前是否待判、答案是否仍隐藏、应填写的决策文件、有效字段和复跑命令。

当前没有输出评审操作清单。

发布标准

Governed 和 Library 至少需要 5 个 output eval cases，并覆盖 file-backed、near-neighbor、boundary case、execution evidence 和 blind A/B review pack。

人工结论

没有 reviewer 决策时只显示 pending；只有真实决策文件会进入一致率和分歧统计。

评审方式

先打开 reports/output_review_kit.html 做盲评，填入 reports/output_review_decisions.json，再用 reports/output_review_adjudication.md 核对答案 key。

运行方式

reports/output_execution_runs.md 会标明 recorded fixture、command run 或 model run；只有 provider runner 返回 model metadata 时才算 model-executed。

运行矩阵

runtime conformance matrix missing

目标编译

目标数
5
通过数
5
Warn Count
0
Block Count
0
失败数
0

上下文

initial load n/a/n/a; deferred n/a/n/a; top deferred none; resource governance unknown; quality density n/a

编译证据

Review reports/compiled_targets.md before packaging to inspect target adapter modes, generated files, preserved semantics, warnings, and unsupported features.

信任报告

security trust report missing

安全边界

高风险 secret、远程 inline execution、缺失依赖策略或无法解释的脚本接口应阻断 governed release。

Python 兼容

python compatibility report missing

解释器边界

CI 和发布审查以 Python 3.11 兼容为底线；本地更高版本允许的新语法不能绕过兼容门禁。

权限批准

permission governance evidence is missing because trust report is missing

批准策略

高权限能力需要 reviewer、scope、reason、expires_at 和 openai/claude/generic 目标端 enforcement 说明。

权限探针

runtime permission probe report is missing

运行边界

runtime permission probe summary missing

组合治理

skill atlas is missing; portfolio-level conflicts are unknown

下一动作

优先处理真实 portfolio 中的 duplicate names、stale skills、owner gaps，再用运营回路判断真实影响。

运营回路

0 metadata events; adoption 0; missed 0; bad-output 0; risk no-data

漂移信号

事件数
0
采用率
0
漏触发
0
Bad Output Count
0
风险带
no-data

人工批准

warning 可以被有边界地接受，但必须写入 reviewer、理由、范围和到期时间；blocker 与 world-class 完成证据不能通过 waiver 变成通过。

批准概况

0 active waivers; 11 warning gates still need reviewer decision

批准台账

Waiver Count
0
Active Count
0
Expired Count
0
Invalid Count
0
覆盖 Gate
0

批准候选

当前没有需要 reviewer 处理的批准候选。

世界证据

这里列出每个 world-class 证据项的当前状态、完成定义、证据来源、隐私约束和下一步；计划、metadata fallback、待评审和本地命令不会被当成完成证据。

当前没有 world-class 证据条目。

提交清单

每张卡片给出模板、提交文件、准备命令、校验命令、收集要求、通过条件和隐私边界；只有真实 provider、真人、原生权限或真实客户端证据通过 intake 后，才进入 ledger review。

当前没有 world-class 证据操作清单。

蓝图覆盖

Skill OS 2.0 blueprint coverage missing

覆盖边界

蓝图覆盖只证明 2.0 模块、建议 PR、脚本、报告和测试在本地闭环；public world-class 仍以 world-class evidence ledger 的真人和外部证据为准。

公开声明

benchmark reproducibility report missing

声明阻断

无公开声明阻断。

世界证据

world-class evidence ledger is optional until governed or public world-class readiness is claimed

证据台账

world-class evidence ledger missing

证据入口

world-class evidence intake missing

入口边界

intake 只校验证据包格式、来源、隐私和反过度声明；只有 ledger 看到真实 provider、真人、原生权限或真实客户端结果后，world-class 才能进入完成审计。

声明守卫

world-class claim guard missing

声明边界

claim guard 扫描 README、docs 和 reports 中的完成态表述；ledger 未 ready 时，任何英文完成断言、true 状态声明或中文完成态都会阻断发布审查。

注册审计

registry audit is optional until team distribution is required

包体元数据

registry package metadata missing

发布路线

promotion decisions are missing; release notes need reviewer confirmation

包体验证

package verification missing

安装模拟

install simulation missing

权限覆盖

安装模拟会读取解压后的 target adapter，并按 declared_capabilities 核对 security/permission_policy.json 中的有效批准、证据、过期时间和目标端 enforcement 说明。