Agent skill

Bfe Rd Workflow

by bfenetworks in bfenetworks/bfe

引导用户在 bfe 代码库中完成一次完整的功能研发流程,包括需求对齐、文档修改、代码实现、集成测试与回归验证. An agent skill from bfenetworks/bfe.

Apache-2.0Auto-check passedDevOps & Cloud

Install Bfe Rd Workflow

skills CLI
$ npx skills add bfenetworks/bfe --skill bfe-rd-workflow -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install bfenetworks/bfe bfe-rd-workflow --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/bfenetworks/bfe.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/bfe-rd-workflow .claude/skills/bfe-rd-workflow && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
bfe-rd-workflow
GitHub stars
6.3k
Token cost
~1.3k tokens
SKILL.md length
327 words
Files
1
Skills in repo
1
Repo updated
First seen
Licence
Apache-2.0

At a glance

引导用户在 bfe 代码库中完成一次完整的功能研发流程,包括需求对齐、文档修改、代码实现、集成测试与回归验证. An agent skill from bfenetworks/bfe.

  • Works in 9 steps: 需求澄清与范围界定 → 修改 modifications 文档 → 更新 configuration 文档 → …
  • Tasks that involve Cloud networking
  • SKILL.md covers 触发语, 执行原则, 研发阶段 and 常见陷阱, plus 1 more section
  • Calls go and git

What it does

Bfe Rd Workflow is an agent skill from bfenetworks/bfe. 引导用户在 bfe 代码库中完成一次完整的功能研发流程,包括需求对齐、文档修改、代码实现、集成测试与回归验证。

Its SKILL.md is about 1.3k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in DevOps & Cloud, covering Cloud networking. It works with Git and Go. The repository describes itself as: A modern layer 7 load balancer from baidu. The licence is Apache-2.0.

When your agent uses it

  • Tasks that involve Cloud networking

Example prompts

  • “/bfe-rd-workflow”

Workflow steps

9 steps, taken from the step headings in SKILL.md.

  1. 需求澄清与范围界定
  2. 修改 modifications 文档
  3. 更新 configuration 文档
  4. 更新 sys_design 文档
  5. 代码实现
  6. 编写集成测试设计文档
  7. 编写集成测试代码
  8. 回归验证
  9. 收尾与总结

What it can do on your machine

Read from SKILL.md and the folder at commit c95df40. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • go
    • git

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Bfe Rd Workflow loads about 1.3k tokens when it runs. Until then it costs about 18 tokens; SKILL.md has 327 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~18
When it runs · the whole SKILL.md, loaded when a task matches
~1.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from bfenetworks/bfe at commit c95df40, republished under its Apache-2.0 licence (© bfenetworks). 327 words, ~1,270 tokens.

Download SKILL.mdSave it as .claude/skills/bfe-rd-workflow/SKILL.md (or your agent's skills folder).
name
bfe-rd-workflow
description
引导用户在 bfe 代码库中完成一次完整的功能研发流程,包括需求对齐、文档修改、代码实现、集成测试与回归验证。

BFE 代码库研发流程

本 Skill 适用于在 bfe/ 目录中新增或修改 BFE 功能(例如 AI Gateway 的多 API-Key 会话亲和性、限流、计费、路由、协议支持等)的完整研发流程。

触发语

当用户提出以下类型请求时启用本流程:

  • “我要实现 xxx 功能”
  • “请帮我完成 xxx 的代码与测试”
  • “请在 BFE 中增加/修改 xxx”

执行原则

  1. 分阶段暂停确认:本流程将研发拆分为多个 Phase。每个 Phase 执行完毕后,必须暂停并等待用户确认“可以继续”后,再进入下一个 Phase。不要在未获得用户确认的情况下自动推进到下一阶段。
  2. Git 提交前确认:在任何 git commit、git push 或其他会改变 Git 仓库状态的操作之前,必须先向用户说明变更内容并取得明确同意。禁止自动执行 Git 提交或推送。
  3. Git push 默认目标为 origin:如果获得用户授权执行 git push,默认推送到 origin 远程仓库,而不是 upstream。除非用户明确指定其他 remote,否则不使用 upstream。

研发阶段

Phase 0. 需求澄清与范围界定
  1. 让用户明确:
    • 功能目标(一句话描述)
    • 验收标准(必须通过的测试/行为)
    • 影响范围(是否改配置协议、是否改 Redis/状态、是否影响既有请求路径)
  2. 判断是否为“非平凡改动”(多文件、有架构选择、用户偏好影响实现):
    • 是 → 调用 EnterPlanMode,先出设计文档/plan 再执行。
    • 否 → 直接进入 Phase 1。

完成本阶段后暂停,等待用户确认后再进入下一阶段。

Phase 1. 修改 modifications 文档

BFE 侧的改动必须在 bfe/docs/zh_cn/modifications/ 留下修改说明,便于后续维护与审计。

  1. 在 bfe/docs/zh_cn/modifications/ 下新建日期化目录,例如:
    bfe/docs/zh_cn/modifications/2026-08-26-ai-key-session-affinity/
  2. 在该目录下创建 design-changes.md,包含:
    • 背景与目标
    • 主要改动点(数据结构、模块行为、新增字段)
    • 配置影响
    • 兼容性说明
  3. 如已有同主题目录,则更新其中的文档,不要重复创建。

完成本阶段后暂停,等待用户确认后再进入下一阶段。

Phase 2. 更新 configuration 文档

如果改动涉及配置文件(cluster_conf.data、mod_*.conf、bfe.conf 等),必须同步更新中英文配置文档:

  1. 中文:bfe/docs/zh_cn/configuration/
  2. 英文:bfe/docs/en_us/configuration/

更新原则:

  • 新增字段要说明类型、默认值、是否必填、示例。
  • 修改字段要说明前后行为差异。
  • 两个语言目录尽量保持内容一致。

完成本阶段后暂停,等待用户确认后再进入下一阶段。

Phase 3. 更新 sys_design 文档

如果改动涉及系统设计、协议交互或状态存储,需要更新 bfe/docs/zh_cn/sys_design/:

  1. 查找现有相关文档(例如 multi_api_key.md)。
  2. 在文档中新增/修改对应章节,说明:
    • 方案选型
    • 数据流
    • 状态存储(如 Redis key 格式、TTL)
    • 边界情况与优化
  3. 如需新增独立文档,直接创建 .md 文件,并在相关文档中建立链接。

完成本阶段后暂停,等待用户确认后再进入下一阶段。

Phase 4. 代码实现
  1. 阅读相关源码,确认:
    • 配置结构体(bfe_config/bfe_cluster_conf/cluster_conf/)
    • 模块加载与处理逻辑(bfe_modules/mod_*)
    • 反向代理与 Key 选择逻辑(bfe_server/reverseproxy.go 及相关测试)
  2. 做最小改动,优先匹配现有代码风格:
    • 不引入未使用的依赖。
    • 不修改与本次需求无关的逻辑。
  3. 关键实现完成后,先跑模块级单元测试:
    bash
    cd bfe
    go test ./bfe_server/... ./bfe_modules/<相关模块>/...

完成本阶段后暂停,等待用户确认后再进入下一阶段。

Phase 5. 编写集成测试设计文档

BFE 的集成测试设计文档位于 bfe/tests/integration/测试设计文档/。在写测试代码之前,先补充对应场景的设计文档,便于后续维护与评审。

  1. 在 bfe/tests/integration/测试设计文档/ 下新建 scenario 目录,命名规范:
    scenario-SC<序号>-<简短中文描述>
  2. 编写 场景说明.md,包含:
    • 场景背景与目的
    • 运行模式
    • 涉及的 BFE 配置文件
    • 测试 Cluster 与关键配置(AIConf、KeyPolicy、路由表等)
    • 测试例列表
    • 每个测试例的详细说明(目的、前置条件、请求构造、执行步骤、预期结果、清理)
    • 公共基础设施
    • 与其他场景的依赖关系
  3. 为每个测试例编写独立的 TC-<序号>-<测试例名称>.md,结构与现有 TC 文档保持一致。
  4. 更新 bfe/tests/integration/测试设计文档/测试场景总体说明.md:
    • 在“场景清单”中新增一行;
    • 在“场景与测试例对应关系”中新增该 scenario 的 TC 列表。

完成本阶段后暂停,等待用户确认后再进入下一阶段。

Show full SKILL.md (141 more words)Show less
Phase 6. 编写集成测试代码

BFE 的集成测试代码位于 bfe/tests/integration/implementation/。

  1. 选择最接近的现有 scenario 作为模板(例如需要 Redis 的可参考 scenario-SC07-ai-rate-limit-redis-key)。
  2. 新建 scenario 目录,命名规范:
    scenario-SC<序号>-<简短英文描述>
  3. 复制模板后,修改 testdata/ 中的:
    • server_data_conf/host_rule.data:host
    • server_data_conf/route_rule.data:cluster
    • cluster_conf/gslb.data:cluster 名称
    • mod_ai_route/ai_route.data:apikey owner / route binding
    • mod_ai_token_auth/mod_ai_token_auth.conf、mod_ai_rate_limit/mod_ai_rate_limit.conf 等
  4. 编写 _test.go,覆盖:
    • 正常路径
    • 功能开关开启/关闭对比
    • 边界场景(单 Key、多 ClientKeyId、BFE 重启后状态保持、错误码/降级)
  5. 运行集成测试:
    bash
    cd bfe
    go test ./tests/integration/implementation/scenario-SC<xx>-<描述>/... -v

完成本阶段后暂停,等待用户确认后再进入下一阶段。

Phase 7. 回归验证
  1. 运行本次新增 scenario 的测试。
  2. 运行与被改动模块相关的既有 scenario:
    • 多 API-Key 轮换:scenario-SC02-multi-api-key
    • Redis 限流:scenario-SC07-ai-rate-limit-redis-key
  3. 运行相关单元测试:
    bash
    go test ./bfe_server/... ./bfe_modules/mod_ai_route/... ./bfe_modules/mod_ai_token_auth/...
  4. 如有失败,优先修复;修复后再次回归,直到全部通过。

完成本阶段后暂停,等待用户确认后再进入下一阶段。

Phase 8. 收尾与总结
  1. 检查是否有注释/文档描述的是旧行为,及时同步更新。

  2. 向用户汇报:

    • 改动了哪些文件
    • 新增/修改了哪些测试
    • 验证结果
    • 仍存在的风险或待决策点(如有)
  3. Git 提交前必须人工确认:如果用户要求或流程需要执行 git commit、git push 等 Git 操作,必须先向用户清晰说明本次提交内容(包含文件清单与主要变更摘要),并取得明确同意后再执行。禁止在未获授权的情况下自动提交或推送。获得授权后,git push 默认推送到 origin,不要推送到 upstream,除非用户明确指定。

常见陷阱

  • cluster 名称不一致:route_rule.data、gslb.data、ai_route 中的 cluster 名称必须一致,否则 BFE 启动会报 no backend conf。
  • ApikeyRouteTableBindings 遗漏:新增 client apiKey 时,必须在 mod_ai_route/ai_route.data 的 ApikeyRouteTableBindings 中为其绑定路由表。
  • Redis 绑定 key 格式:确认代码中实际使用的 key 格式(如 bfe:ai:key_affinity:<cluster>:<client_key_id>)。
  • 单 Key 优化:如果实现了“单 Key 不走 Redis”,需要同时避免读和写,否则集成测试里 Exists 会误判。
  • BFE 端口冲突:集成测试里 processEnv.StartBFE 会自动分配端口,不要手动写死。

推荐命令速查

bash
# 模块单元测试
cd bfe
go test ./bfe_server/... ./bfe_modules/mod_ai_route/... ./bfe_modules/mod_ai_token_auth/...

# 单个集成测试场景(详细日志)
go test ./tests/integration/implementation/scenario-SC<xx>-<描述>/... -v

# 相关场景回归
go test ./tests/integration/implementation/scenario-SC02-multi-api-key/...
go test ./tests/integration/implementation/scenario-SC07-ai-rate-limit-redis-key/...

© bfenetworks, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/bfe-rd-workflow of bfenetworks/bfe.

Open the folder on GitHubat commit c95df40

Compare with similar skills

Bfe Rd Workflow next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Bfe Rd Workflow compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Bfe Rd Workflow this skillbfenetworks/bfe6.3k—~1.3kAutomated safety check: PassApache-2.0
NGINX Ingress Controller Feature Checklistsnginx/kubernetes-ingress5.1k—~1.4kAutomated safety check: PassApache-2.0
NGINX Ingress Policy CRD Guidenginx/kubernetes-ingress5.1k—~2kAutomated safety check: PassApache-2.0
Detecting Directory Listingjeremylongshore/tons-of-skills-marketplace2.8k—~1.8kAutomated safety check: NotesMIT
Network TroubleshootingMindrally/skills268—~2.2kAutomated safety check: PassApache-2.0
Go-Redis Release Preparationredis/go-redis22k—~1.1kAutomated safety check: PassBSD-2-Clause

Similar skills

  • Gives step-by-step checklists for adding Ingress annotations, VirtualServer fields and Helm values to the NGINX Kubernetes Ingress Controller, with common gotchas.

    5.1k GitHub stars~1.4k tokensUpdated today
    DevOps & CloudAuto-check passed
  • NGINX Ingress Policy CRD Guide

    nginx/kubernetes-ingress

    Step-by-step checklist for adding a new Policy CRD type to the NGINX Ingress Controller, from the Go types and validation to config generation and templates.

    5.1k GitHub stars~2k tokensUpdated today
    DevOps & CloudAuto-check passed
  • Detecting Directory Listing

    jeremylongshore/tons-of-skills-marketplace

    Probe a target for directories that return auto-generated index listings instead of denying or serving a specific file — exposes the full file tree under any reachable directory, including files the…

    2.8k GitHub stars~1.8k tokensUpdated today
    DevOps & CloudAuto-check: notes
  • Network Troubleshooting

    Mindrally/skills

    Systematic, safety-first diagnosis of developer network failures such as connection refused, DNS resolution errors, TLS certificate problems, proxy failures, and package registry timeouts.

    268 GitHub stars~2.2k tokensUpdated 1 mo ago
    DevOps & CloudAuto-check passed
  • Official

    Prepares a go-redis release locally: picks the next semver, gathers merged PRs, writes the RELEASE-NOTES entry and bumps versions, without publishing.

    22k GitHub stars~1.1k tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Syntax reference for KFL2, the CEL-based display filter language used to search Kubernetes network traffic captured by Kubeshark, loaded before any filter is written.

    12k GitHub stars~3.6k tokensUpdated yesterday
    DevOps & CloudAuto-check passed

Works with

Categories

Questions about Bfe Rd Workflow

What does Bfe Rd Workflow do?

引导用户在 bfe 代码库中完成一次完整的功能研发流程,包括需求对齐、文档修改、代码实现、集成测试与回归验证. An agent skill from bfenetworks/bfe. Bfe Rd Workflow is an agent skill from bfenetworks/bfe.

When should I use Bfe Rd Workflow?

Bfe Rd Workflow fits situations like: tasks that involve Cloud networking.

How do I install Bfe Rd Workflow in Claude Code?

Run `npx skills add bfenetworks/bfe --skill bfe-rd-workflow -a claude-code`. Or copy the skill folder (skills/bfe-rd-workflow in bfenetworks/bfe) into .claude/skills/bfe-rd-workflow in your project. Claude Code loads it when a task matches its description.

How do I install Bfe Rd Workflow in Codex?

Run `npx skills add bfenetworks/bfe --skill bfe-rd-workflow -a codex`. Or copy the skill folder (skills/bfe-rd-workflow in bfenetworks/bfe) into .agents/skills/bfe-rd-workflow in your project. Codex loads it when a task matches its description.

Can I use Bfe Rd Workflow in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add bfenetworks/bfe --skill bfe-rd-workflow -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/bfe-rd-workflow, .gemini/skills/bfe-rd-workflow, .github/skills/bfe-rd-workflow and .opencode/skills/bfe-rd-workflow in your project.

What does Bfe Rd Workflow need to run?

Going by SKILL.md and its folder, Bfe Rd Workflow needs the command-line tools its instructions call (go and git).

Does Bfe Rd Workflow access the network?

SKILL.md contains no URLs. Its commands use git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Bfe Rd Workflow safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Bfe Rd Workflow use?

Bfe Rd Workflow is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Bfe Rd Workflow use?

About 1.3k tokens (SKILL.md is roughly 5.1k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Bfe Rd Workflow?

Skills that share tags, products or a category with Bfe Rd Workflow: NGINX Ingress Controller Feature Checklists (nginx/kubernetes-ingress, 5.1k stars), NGINX Ingress Policy CRD Guide (nginx/kubernetes-ingress, 5.1k stars), Detecting Directory Listing (jeremylongshore/tons-of-skills-marketplace, 2.8k stars) and Network Troubleshooting (Mindrally/skills, 268 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Bfe Rd Workflow?

bfenetworks (a GitHub organization) maintains it in bfenetworks/bfe, which has 6,268 GitHub stars. The repository was last updated on October 8, 2026.

Source: bfenetworks/bfe on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.