Token导航 LogoToken导航TokenDH.com
研究检索执行命令github未标认证来源可访问许可证需确认审计提醒

flow-governance-review流程治理审查

Agent Skill

flow-governance-review 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

216

周安装

9

GitHub Stars

4

下载量

72
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:flow-governance-review(流程治理审查)
来源仓库:https://github.com/tiangong-lca/skills
仓库路径:skills/flow-governance-review
安装命令:
npx skills add https://github.com/tiangong-lca/skills --skill flow-governance-review
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/tiangong-lca/skills --skill flow-governance-review

简介

用于查找、检索和筛选相关信息。flow-governance-review 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

  • 适合根据关键词快速定位候选结果。
  • 可结合来源仓库继续核验具体用法。
  • 安装前建议确认权限和维护状态。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。
  • 注意是否会触发联网或文件读写。

SKILL.md

Flow Governance Review

Keep local JSON or JSONL payloads as the system of record. This skill is a thin wrapper around the supported CLI governance commands.

Do not use this skill for:

  • arbitrary remote CRUD outside the explicit CLI commit commands
  • hidden OpenClaw orchestration
  • private Python remediation helpers
  • remote scope export

Runtime Model

  • The canonical entrypoint is node scripts/run-flow-governance-review.mjs <command>....
  • Write outputs to an explicit directory such as /abs/path/artifacts/<case_slug>/....
  • Supported commands are all CLI-backed:

- review-flows -> tiangong review flow - flow-get -> tiangong flow get - flow-list -> tiangong flow list - materialize-db-flows -> tiangong flow fetch-rows - materialize-approved-decisions -> tiangong flow materialize-decisions - remediate-flows -> tiangong flow remediate - publish-version -> tiangong flow publish-version - publish-reviewed-data -> tiangong flow publish-reviewed-data - build-flow-alias-map -> tiangong flow build-alias-map - scan-process-flow-refs -> tiangong flow scan-process-flow-refs - plan-process-flow-repairs -> tiangong flow plan-process-flow-repairs - apply-process-flow-repairs -> tiangong flow apply-process-flow-repairs - regen-product -> tiangong flow regen-product - validate-processes -> tiangong flow validate-processes

  • publish-reviewed-data is fully CLI-owned for both local preparation and commit-time process publish.
  • There is no Python fallback path and no shell compatibility shim.

Commands

  • flow-get
  • flow-list
  • materialize-db-flows
  • materialize-approved-decisions
  • remediate-flows
  • publish-version
  • publish-reviewed-data
  • build-flow-alias-map
  • scan-process-flow-refs
  • plan-process-flow-repairs
  • apply-process-flow-repairs
  • regen-product
  • validate-processes
  • review-flows

Run them through:

node scripts/run-flow-governance-review.mjs <command> ...

For CLI-backed deterministic governance slices, prefer:

node scripts/run-flow-governance-review.mjs materialize-db-flows \
  --refs-file /abs/path/flow-refs.json \
  --out-dir /abs/path/materialized \
  --fail-on-missing

node scripts/run-flow-governance-review.mjs materialize-approved-decisions \
  --decision-file /abs/path/approved-decisions.json \
  --flow-rows-file /abs/path/materialized/review-input-rows.jsonl \
  --out-dir /abs/path/decision-artifacts

node scripts/run-flow-governance-review.mjs review-flows \
  --rows-file /abs/path/flows.jsonl \
  --out-dir /abs/path/review

node scripts/run-flow-governance-review.mjs remediate-flows \
  --input-file /abs/path/invalid-flows.jsonl \
  --out-dir /abs/path/remediation

node scripts/run-flow-governance-review.mjs publish-version \
  --input-file /abs/path/ready-flows.jsonl \
  --out-dir /abs/path/publish \
  --dry-run

node scripts/run-flow-governance-review.mjs publish-reviewed-data \
  --flow-rows-file /abs/path/reviewed-flows.jsonl \
  --original-flow-rows-file /abs/path/original-flows.jsonl \
  --out-dir /abs/path/publish-reviewed

node scripts/run-flow-governance-review.mjs build-flow-alias-map \
  --old-flow-file /abs/path/old-flows.jsonl \
  --new-flow-file /abs/path/new-flows.jsonl \
  --out-dir /abs/path/alias-map

node scripts/run-flow-governance-review.mjs scan-process-flow-refs \
  --processes-file /abs/path/processes.jsonl \
  --scope-flow-file /abs/path/flows.jsonl \
  --out-dir /abs/path/scan

node scripts/run-flow-governance-review.mjs plan-process-flow-repairs \
  --processes-file /abs/path/processes.jsonl \
  --scope-flow-file /abs/path/flows.jsonl \
  --scan-findings /abs/path/scan/scan-findings.json \
  --out-dir /abs/path/repair-plan

node scripts/run-flow-governance-review.mjs apply-process-flow-repairs \
  --processes-file /abs/path/processes.jsonl \
  --scope-flow-file /abs/path/flows.jsonl \
  --scan-findings /abs/path/scan/scan-findings.json \
  --out-dir /abs/path/repair-apply

node scripts/run-flow-governance-review.mjs regen-product \
  --processes-file /abs/path/processes.jsonl \
  --scope-flow-file /abs/path/flows.jsonl \
  --out-dir /abs/path/regen \
  --apply

node scripts/run-flow-governance-review.mjs validate-processes \
  --original-processes-file /abs/path/before.jsonl \
  --patched-processes-file /abs/path/after.jsonl \
  --scope-flow-file /abs/path/flows.jsonl \
  --out-dir /abs/path/validate

Not Supported

The following legacy commands were intentionally removed with the Python runtime:

  • openclaw-entry
  • openclaw-full-run
  • run-governance
  • flow-dedup-candidates
  • export-openclaw-*
  • apply-openclaw-*
  • validate-openclaw-*

If you need one of those workflows, add it first as a native tiangong review... or tiangong flow... command instead of rebuilding it inside this skill.

Preferred Usage

Use the supported commands as composable slices:

  1. materialize-db-flows when the task must bind to real DB rows
  2. review-flows
  3. materialize-approved-decisions after merge decisions are approved
  4. remediate-flows
  5. build-flow-alias-map when version cleanup produced old/new scopes
  6. scan-process-flow-refs
  7. plan-process-flow-repairs
  8. apply-process-flow-repairs
  9. validate-processes
  10. publish-version or publish-reviewed-data

Standard Outputs

  • flow-alias-map.json when alias building is applicable
  • scan-findings.json and repair-summary.json when process snapshots are provided
  • publish-report.json from publish-reviewed-data
  • prepared-flow-rows.json and flow-version-map.json from publish-reviewed-data
  • skipped-unchanged-flow-rows.json from publish-reviewed-data when --original-flow-rows-file is provided
  • resolved-flow-rows.jsonl, review-input-rows.jsonl, and fetch-summary.json from materialize-db-flows
  • flow-dedup-canonical-map.json, flow-dedup-rewrite-plan.json, manual-semantic-merge-seed.current.json, and blocked-clusters.json from materialize-approved-decisions

Example Output Layout

Write generated machine outputs to an explicit directory outside the skill source tree. Typical bundles include:

  • /abs/path/artifacts/<case_slug>/flow-processing/datasets/: shared flow pool, invalid-input scope, resolved flow pool, reusable process_pool.jsonl
  • /abs/path/artifacts/<case_slug>/flow-processing/validation/: grouped validation failures that still matter for remediation planning
  • /abs/path/artifacts/<case_slug>/flow-processing/naming/remaining-after-aggressive/: post-aggressive completeness summaries and zero-process residuals
  • /abs/path/artifacts/<case_slug>/flow-processing/naming/zero-process-completion-pack/: reference review materials retained for follow-up
  • /abs/path/artifacts/<case_slug>/flow-processing/remediation/: deterministic remediation and publish preparation artifacts
  • /abs/path/artifacts/<case_slug>/flow-remediation-batch-smoketest/: historical smoke-test evidence for remediation-helper startup checks

Load References On Demand

  • references/workflow.md: command matrix, outputs, removed surface, and recommended sequencing.
  • references/env.md: canonical CLI env expectations for read, review, and publish commands.
  • references/real-db-first-runbook.md: real-DB-first execution guardrails, refs-file shape, and blocked-case handling.
  • references/decision-schema.md: approved decision file schema, merge examples, and downstream artifact meanings.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

34.9%
按下载量换算25

Claude

26.99%
按下载量换算19

Cursor

19.55%
按下载量换算14

Gemini CLI

8.84%
按下载量换算6

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

执行命令

安装流程涉及命令执行,可能通过 npx skills add https://github.com/tiangong-lca/skills --skill flow-governance-review 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills