systematic-debugging skill A
systematic-debugging is agent-read markdown (skill) from minleev5/devkeel: 系统化诊断和修复 Bug、测试或构建失败、异常行为、集成故障与性能回归。先建立可判定的反馈信号、追踪根因并逐一验证可证伪假设;仅在现有错误、测试、日志、diff、debugger 或 profiler 无法区分假设,且临时写入已获授权时使用定向 HTTP 插桩。.
Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.
What the file says # Systematic Debugging
以证据闭环解决问题。插桩是遇到运行时观测缺口后的升级手段,不是默认入口。
## 权限边界
- 用户只要求分析、诊断或解释时保持只读;可以复现、检查和运行非破坏性诊断,但不得修复或向
源码写入临时探针。
- 用户明确要求修复时,只修改已授权目标。临时插桩同样是源码写入,必须处于该授权范围内。
- 无法从当前环境复现时,列出已尝试的方法,并请求最小必要的日志、trace、请求、录屏或环境访问;
不用猜测代替证据。
## 调试状态机
按顺序推进,但让简单问题走快速路径:
1. **定义症状**:写清输入、期望、实际行为和首次已知异常位置。
2. **建立反馈信号**:找到一条能命中真实路径、对该症状判红/判绿、可由 Agent 重复运行的命令或
现场证据。已有失败测试或确定性错误就是现成信号,不另造 harness。
3. **复现并最小化**:确认是用户报告的同一个问题;逐项减少输入、配置、调用方和环境差异。
4. **追踪根因**:完整读取错误与 stack,检查近期变化、正常样例、依赖和数据来源,向上追到错误
状态首次产生的位置。
5. **形成假设**:生成少量有排序、可证伪的候选;每个候选写出若成立应观察到什么。一次只测试
一个假设和一个变量。
6. **修复与验证**:只有根因得到证据支持且用户已授权修复时,才在正确接缝添加回归测试并实施
单一修复;重新运行最小复现、原始反馈信号和邻近测试。
简单问题可以快速完成这些步骤,但不能省略症状、证据、根因和验证。需要更完整的反馈循环、
最小化、数据回溯、性能分支或失败升级策略时,读取 `references/diagnosis.md`。
## 插桩升级门禁
只有以下条件全部满足时才读取 `references/instrumentation.md` 并启动 collector:
1. 已捕获精确症状,且有可重复反馈信号或可靠现场材料;
2. 剩余不确定性来自不可见的运行时状态、时序或跨边界数据流;
3. stack、测试缩小、diff/bisect、现有日志、debugger/REPL 或 profiler 仍不能区分假设;
4. 每个计划探针对应一个明确预测,并且临时代码写入已获授权。
禁止“记录一切再搜索”。性能回归默认使用基线、profiler、query plan 或 bisect;只有需要验证某个
具体运行时预测时才插桩。
## 停止与回退
- 一个假设被否定后记录新证据,回到根因追踪,不在失败改动上叠加下一次试修。
- 连续三次修复尝试失败,停止尝试第四次,检查共享状态、耦合和测试接缝是否暴露架构问题,并让
用户决定是否扩大到架构调整。
… Read the whole file at its exact version.
How to install
Latest version
mdr add minleev5/devkeel/systematic-debugging@v1.0.2Copy
Exact content
mdr add minleev5/devkeel/systematic-debugging@sha256:c7717dfb4bc234ffCopy
Pin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.
Badge
[](https://markdownregistry.com/a/art_m3bxjadtiassokll) Copy 1 badge views in 30 days
Versions
Audit of the latest version A 17 of 17 checks passed. Deterministic, no model, same answer every run.
pass: Frontmatter block present pass: Frontmatter declares a name pass: Frontmatter declares a description pass: Size between 200 bytes and 200 KB (4120 bytes) pass: No zero-width or bidi control characters pass: No instruction hidden inside an HTML comment pass: No link to an exfiltration or paste host pass: No credential-shaped string pass: No instruction to send local credentials anywhere pass: No text hidden with inline styles pass: No prompt-injection phrasing pass: No curl or wget piped into a shell pass: No recursive delete of root, home or parent pass: No instruction to read or print local credentials pass: No base64 blob over 200 characters pass: No link to a raw IP address pass: No script tag
Source GitHub
minleev5/devkeel · 3 stars · license MIT · pushed 2026-09-24 · branch master
API GET https://markdownregistry.com/api/v1/artifacts/art_m3bxjadtiassokll
GET https://markdownregistry.com/api/v1/resolve?ref=minleev5/devkeel/systematic-debugging
GET https://markdownregistry.com/api/v1/blob/c7717dfb4bc234ff8096a76f5490c410dd1d0fa28ba464d7a0ace4d66e2a7b2bCopy
Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
More from minleev5/devkeel brainstorming skill
minleev5/devkeel · .harness/skills/brainstorming/SKILL.md · 探索、比较、挑战或补全开发方案;按 Lite/Full 深度进行单题访谈,并在用户同意持久化后维护 OpenSpec Living brainstorm。
v9.1.1 · audit A · 3 stars
changelog skill
minleev5/devkeel · .harness/skills/changelog/SKILL.md · Use when 编辑 changelog、发布 CLI 或模板版本、或判断某项用户可见变更是否需要进入版本记录。
v3.0.0 · audit A · 3 stars
commit skill
minleev5/devkeel · .harness/skills/commit/SKILL.md · 用户明确选择 commit、push 或 MR/PR 后使用的原子交付流程。任务完成或归档本身不触发;在授权范围内完成交付目标及其必要动作。
v1.2.4 · audit A · 3 stars
doc-maintenance skill
minleev5/devkeel · .harness/skills/doc-maintenance/SKILL.md · CLI 文档维护与一致性检查流程
v1.2.0 · audit A · 3 stars
domain-init skill
minleev5/devkeel · .harness/skills/domain-init/SKILL.md · 对项目进行全盘扫描,识别领域类型和技术栈,通过多 subAgent 并行扫描多维度, 生成项目知识 docs 与专属 rules、skills 和 agents。使用「自动 baseline + 交互式增强」…
v2.0.2 · audit A · 3 stars
health-check skill
minleev5/devkeel · .harness/skills/health-check/SKILL.md · 一键检查项目代码健康度指标
v1.0.0 · audit A · 3 stars
instruction-maintenance skill
minleev5/devkeel · .harness/skills/instruction-maintenance/SKILL.md · 精简并校验 AGENTS.md、OpenSpec schema/instruction、OPSX commands、SKILL.md、prompt 与 artifact 模板。用于审查当前 diff…
v1.0.0 · audit A · 3 stars
jira-defect-orchestrator skill
minleev5/devkeel · .harness/skills/jira-defect-orchestrator/SKILL.md · 独立编排 Jira 缺陷的取证、修复、验证、回填与 MR 交付,按决策风险确认。适用于用户提供 Jira URL 或 issue key,要求结合运行日志、配置、只读业务数据和代码证据处理生产或测试环境缺陷;普通无 Jira…
v1.3.2 · audit A · 3 stars
new-subcommand skill
minleev5/devkeel · .harness/skills/new-subcommand/SKILL.md · 为 DevKeel CLI 新增子命令的标准开发流程
v1.0.0 · audit A · 3 stars
openspec-apply-change skill
minleev5/devkeel · .harness/skills/openspec-apply-change/SKILL.md · 按 OpenSpec 动态 apply instruction 实施 change tasks,并处理 DevKeel 收尾门禁。
v3.2 · audit A · 3 stars
openspec-archive-change skill
minleev5/devkeel · .harness/skills/openspec-archive-change/SKILL.md · 检查完成状态、评估 spec sync,并按 schema 收尾门禁归档 OpenSpec change。
v3.3 · audit A · 3 stars
openspec-bulk-archive-change skill
minleev5/devkeel · .harness/skills/openspec-bulk-archive-change/SKILL.md · 批量检查 spec 冲突,并按 schema 门禁归档用户明确选择的多个 OpenSpec changes。
v2.2 · audit A · 3 stars
Every file in minleev5/devkeel
Other files named systematic-debugging systematic-debugging skill
obra/superpowers · skills/systematic-debugging/SKILL.md · Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes
git:20260724.c74782e · audit A · 291,204 stars
systematic-debugging skill
jnmetacode/superpowers-zh · skills/systematic-debugging/SKILL.md · 遇到任何 bug、测试失败或异常行为时使用,在提出修复方案之前执行
v1.0.0 · audit A · 8,203 stars
systematic-debugging skill
a5c-ai/babysitter · library/methodologies/superpowers/skills/systematic-debugging/SKILL.md · Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes. Requires root cause…
git:20260601.da7723a · audit A · 1,809 stars
systematic-debugging skill
a5c-ai/babysitter · library/methodologies/rpikit/skills/systematic-debugging/SKILL.md · Structured debugging methodology using hypothesis-driven investigation, log analysis, and bisection to isolate and…
git:20260605.09bdbf7 · audit A · 1,809 stars
systematic-debugging skill
ganyuanran/aegis · skills/systematic-debugging/SKILL.md · Use when encountering a bug, test failure, or unexpected behavior, before proposing fixes
git:20260907.f725a76 · audit A · 1,280 stars
systematic-debugging skill
mateaix/mateclaw · mateclaw-server/src/main/resources/skills/systematic-debugging/SKILL.md · 4-phase root cause debugging: understand bugs before fixing.
v1.1.0 · audit A · 1,127 stars
systematic-debugging skill
sethgammon/citadel · skills/systematic-debugging/SKILL.md · 4-phase root cause analysis: observe, hypothesize, verify, fix. Enforces investigation before any code changes…
git:20260624.445e974 · audit A · 923 stars
systematic-debugging skill
taracodlabs/aiden · skills/systematic-debugging/SKILL.md · Four-phase root cause investigation for bugs and unexpected behavior
v1.0.0 · audit A · 843 stars
Browse by kind , by grade A , or by owner .