TypingMindTypingMindThe best frontend for AI
Analyze Default logo

Analyze Default

Organization
devcodex-labs
analyze-default

默认分析工作流规范 — 只读多轮分析、代码事实优先、analyze-lite 关联联查与 PCV 收敛验证

Overview

Publisherdevcodex-labs
Repositorydevcodex
Skill nameanalyze-default
Stars
433
Forks
44
Bundled files
1
LicenseAGPL-3.0
Links
Skill sourceRepositoryHomepage
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by devcodex-labs on GitHub. Read the source before you install it.

Installation

Install the Analyze Default AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/devcodex-labs/devcodex/tree/main/content/skills/analyze-default
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/devcodex-labs/devcodex.git /tmp/devcodex
mkdir -p .claude/skills
cp -r /tmp/devcodex/content/skills/analyze-default .claude/skills/analyze-default
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Analyze Default in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Analyze Default on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Analyze Default is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Analyze Default Skill

定位

analyze-default 承接 analyze.default 的执行细节。instructions/13-analyze.instructions.md 只保留工作流入口、只读底线和路由索引；默认分析的轮次、证据、收敛和输出字段由本 Skill 负责。

触发条件

  • 用户要求分析、判断、解释、定位原因、评估是否合理，但当前目标是结论而不是直接改文件。
  • 语义初判为 analyze，且不满足 analyze.research 的技术调研、选型或外部资料优先条件。
  • 用户要求“全面审查 / 多维度审查 / review / audit”时不得进入本 Skill，应路由到 audit。

只读边界

  • analyze 是项目内容只读工作流，禁止修改源码、规范与配置；只允许写分析报告、记忆，以及经 spec-governance 语义分流后必须写入的 active-root 运行态台账。
  • 发现必须修改的问题时，只能在结论里建议切换 dev 或 fix，不得在 analyze 内直接实施。
  • 用户先给结论、根因或方案假设时，必须独立取证；核验成立才采纳，并说明证据。

执行流程

步骤要求
A1 问题界定写一句话分析目标、边界、输入证据和不分析范围
A1a 规模路由broad scan 前调用 skill-gap-analysis 的 ProjectArtifactScaleRoutingGate；先识别项目并形成 ScaleDecisionRecord，再决定 single-pass / batched / sampled+deep-read / blocked
A1b 增量与精度当 decision 为 batched / sampled+deep-read，或用户要求增量/逐文件/完整深度时，必须加载 incremental-project-analysis：冻结 depthTier、尝试加载/创建 ProjectKnowledgeSnapshot、按 SelectiveInvalidation 生成本轮 plan；禁止只靠自然语言摘要复用
A2 事实取证先查真实代码、文档、配置或运行证据，再对照计划或用户说法
A3 多轮分析至少 3 轮；连续 2 轮无新发现后才可收敛
A4 analyze-lite CRS建立关联文件集合，收敛前反向联查是否遗漏关键消费者
A5 PCV 汇总验证对每条结论执行去重、实证核查、三列验证、分级和推荐结论
A5-ef EvidenceFreshness对“已验证 / 推荐 / 可确认 / 完整”等 strong claim 生成或引用 ClaimEvidenceIndexV1 与 StaleEvidenceLintDecisionV1；summary-only 只能导航，不能支撑最终强结论
A5a 治理评估完成结论合理性评估后，对当前中性候选执行 PostAssessmentGovernanceIntakeGate；复合意图逐项落账，record.none 提供 challenge evidence
A6 报告输出结论必须包含合理性、可实施性、收益、验证状态和影响范围；non-small 须 Theme+Detail 双产物；确认清单须 CoverageMatrix
A6b 分批交付若走 incremental-project-analysis 分批：每 accepted 批输出 BatchProgressCard；全批后 GlobalOptimizationBacklog + 双层 ValidationResult

证据门禁

CodeTruthFirstGate

当结论依赖代码是否存在、如何实现、路径是否正确或行为是否真实时：

  1. 先用关键词、接口路径、类名、方法名或配置名探索真实代码。
  2. 再与需求、计划或报告中的路径对比。
  3. 结论标为 按计划实现、架构偏差但功能完整、未实现 或 证据不足。

禁止只对计划文档里的路径执行存在性检查后直接判定未实现。

ControlPlaneAdviceInventoryGate（建议路径 inventory · PI-176 / PF-181 / PF-178 同簇）

当结论是设计建议、分级方案、体系拆分、Gate/复审强度/控制面如何改，或用户问「能否/应该如何」且答案依赖 DevCodex 或当前仓库如何实现 时，在输出推荐矩阵或「最优方案」之前必须：

  1. 对 source-root（规范仓如 devcodex，非仅 user-global 镜像）做 ExistingCapabilityInventory：既有 Gate 名、Skill Owner、runtime（如 selectReviewClass）、探针、相关 PF/PI、文案漂移面。
  2. 输出最小 inventory 表或等价证据：reusePoint / currentBehavior / gap / 禁平行声明（不得静默发明平行等级名或新 Gate 哲学体系）。
  3. 仅对话记忆、user-global Skill 镜像或未 inventory 时，禁止写「已验证最优 / 项目已支持 / 可直接按此实施」。

触发面示例：C19/R 档、ECR、复审清单、Hook/CLI/MCP、validate、host projection、规范吸纳。
低风险且与项目实现无关的纯概念解释可写 N/A + skipReason，不得把控制面建议降级为 N/A。
与 expert-output-quality 的 CodeTruthEvidenceMatrixGate / SolutionFitAgainstRepoGate 同向；chat 路径同样适用，不因未进 CP1 而豁免。

ProfileTruthReconciliationGate（targeted）

项目级 analyze 在 A2 事实取证时必须调用 load-profile 的 ProfileTruthReconciliationGate targeted 模式：先把 Profile 声明当作待核对输入，再用当前代码、配置、package、运行证据和正式需求建立 ProfileTruthMatrix。若出现 stale-profile 或 unverifiable，本轮结论必须采用可验证事实并明确可信度；若出现 stale-code-or-doc 或 intentional-exception，必须说明目标态/政策依据和消费者影响。

analyze 只矫正结论，不修改 Profile。需要修订 Profile 时在 upgradeAdvice 指向独立 dev/fix/self-fix。低风险单文件且结论与项目事实无关时可写 N/A + skipReason，不得把普通项目级根因分析降级为 N/A。

AnalyzeLiteCRSGate

每轮分析需要维护关联文件集合：

  • seedEvidence：用户输入、截图、报告、路径或初始线索。
  • actualSources：本轮实际读取的代码、文档、配置、测试或日志。
  • relatedConsumers：可能消费该事实的 README、website、Profile、prompts、validate、部署副本或运行时。
  • missingSurface：尚未读取但可能影响结论的关联面。

收敛前必须复查 missingSurface，并说明未继续读取的 skipReason。

ProjectArtifactScaleRoutingGate

项目级 analyze、全库关联联查或用户提示“大目录/文件很多”时必须触发。未形成 ScaleDecisionRecord 前，只允许带排除边界的 bounded inventory；非 single-pass 必须记录 batch budget、checkpoint、timeout/retry 和 invalid-run 排除。抽样深读只能声明“全量 inventory + 代表性深读”，不得宣称逐字全读。

QuestionEvidenceGate

当问题本质是“是否应该 / 哪个更好 / 有没有更好建议 / 推荐什么”时，先判断是否需要 ComparativeResearchGate：

  • 涉及高成本、技术路线、外部平台、产品路线或长期维护时，升级 analyze.research 或补足对比证据。
  • 纯解释、低风险本地事实核验或用户明确要求快速答复时，写 N/A + skipReason。

ReviewFindingIntakeGate

外部审查报告、AI review finding、audit issue 或代码评审发现只能作为线索。每条 finding 必须本地复核，并分类为：

  • must-fix
  • user-decision-required
  • docs-implementation-drift
  • test-coverage-gap
  • already-fixed-or-not-reproduced
  • intentional-design-accepted

不得直接按审查报告文字验证通过；涉及设计如此、兼容策略或产品取舍时，必须记录依据和消费者影响。

GovernanceGateRegistryRef

分析发现规范吸纳、长清单残留、复审遗漏、用户文档、发布门禁、前端缓存或自我进化控制面问题时，不在本 Skill 内复制 Gate 长清单；应引用 spec-governance 的 GovernanceGateRegistry，输出 gateGroup、ownerSkill、trigger、evidence 与 validationRoute。

PCV 收敛验证

PCV动作
PCV-1汇总并去重所有轮次发现
PCV-2对每条结论重新读取对应事实源；运行时数字优先本轮实际执行，不能用记忆历史数字冒充已验证
PCV-2aMeasuredVerificationStandard：探针/validate/测试结论标 已验证 前必须跑生产入口（如 node scripts/test-spec-governance.js、npm run test:core）；隔离 harness 未复用 createCanonicalAwareReader 与 validate 上下文时只能标非权威实验，不得写成 V# 红/绿
PCV-2bEvidenceFreshnessGate：最终结论、推荐方案、覆盖声明和外部 finding 采纳声明必须有 fresh evidence refs；若 StaleEvidenceLintDecisionV1.status=WARN/UNVERIFIED/BLOCK，输出降级措辞或转入 audit/dev/fix，不得继续写强结论
PCV-3补齐合理性、可实施性、收益
PCV-4多路径时给出推荐结论和推荐理由；无后续动作时写 推荐：无后续动作
PCV-5标注 已验证 / 待验证 / 排除；用户可见摘要分列命令、exitCode、权威/实验
PCV-6输出过滤后的最终结论，排除项说明原因

输出最小字段

字段要求
analysisTarget一句话问题定义
scaleDecisionProjectArtifactScaleRoutingGate、六项规模指标、四态决策、预算/checkpoint；N/A 仅限明确单文件且说明理由
rounds至少 3 轮，记录每轮新增发现数
evidenceMap结论到文件、命令或事实源的映射
profileTruthmode、profileTrustState、ProfileTruthMatrix；N/A 时写 skipReason
pcvPCV-1~PCV-6 结果
evidenceFreshnessClaimEvidenceIndexV1.indexDigest、StaleEvidenceLintDecisionV1.status/mode、downgrade/rerun 计数；不触发时写 N/A + skipReason=no-strong-claims
recommendation推荐结论、推荐理由或无后续动作
upgradeAdvice是否建议切换 audit/dev/fix/research，含理由
governanceIntakecandidate ID、评估结论、泛化范围、现有规范状态、复合 record intents、target ledgers、write requirement/evidence、verification state 或 none challenge
coverageMatrix大库/多报告确认清单时：FindingThemeCoverageMatrix（source→mapped/residual/EX）；禁止主题合并冒充零遗漏（ABS-17）
timing长分析：SessionTimingCard 或 N/A+skipReason（ABS-18）

分析交付双产物（ABS-10）

大库或多批分析完成时，除主题优先级清单外，须链接去噪后的 high/medium 明细（或 FindingDetailLedger）；不得用 8 条主题冒充「仅有 8 个问题」。精度档位与宣称边界见 incremental-project-analysis（待建）/ PI-115。

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Analyze Default AI skill do?

默认分析工作流规范 — 只读多轮分析、代码事实优先、analyze-lite 关联联查与 PCV 收敛验证

Why use Analyze Default on TypingMind?

Because you install it once and use it with any model. Analyze Default is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Analyze Default in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/devcodex-labs/devcodex/tree/main/content/skills/analyze-default. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Analyze Default?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Analyze Default?

As many as you like. As long as a model supports skills, you can use Analyze Default with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Analyze Default AI skill free?

Yes. It is published on GitHub by devcodex-labs under the AGPL-3.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

Related AI Skills

View all
openai logo

Openai Docs

openai
OrganizationPopular

Use for Codex models/pricing, scheduled tasks, skills, settings, setup, troubleshooting, customization, automations, and self-knowledge—including 'you,' 'your,' 'this app,' or 'this coding agent' when…

19.9K
127.4K
15 files
View
JuliusBrussee logo

Caveman Explore

JuliusBrussee
CommunityPopular

Read-only repository explorer for cold-start orientation, broad cross-file localization, or when a direct search failed. Skip it when the exact file or symbol is already named. Returns path:line…

6.2K
106.3K
2 files
View
paperclipai logo

Delegate Work

paperclipai
OrganizationPopular

Delegate user-requested work to an existing Paperclip agent, with a durable task reference and clear execution expectations.

16.9K
100.4K
Instructions
View
nexu-io logo

Html Ppt Zhangzara Vellum

nexu-io
OrganizationPopular

A humanities lecture: how Renaissance linear perspective reshaped early cartography — sources, argument, and evidence. Built as a decision-grade academic research deck for faculty, graduate seminar.

11.5K
98.9K
2 files
View

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇