Impact Trace logo

Impact Trace

OrganizationPopular
Prismer-AI
impact-trace

For a change point (function / type / endpoint / SQL column), enumerate its real blast radius across six dimensions with a grep citation per row, then derive the regression test scope. Every impact-table row must trace to a real search hit — never a self-asserted claim.

Overview

PublisherPrismer-AI
RepositoryPrismerCloud
Skill nameimpact-trace
Stars
1.6K
Forks
17
Bundled files
1
LicenseMIT
Links
  • 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 Prismer-AI on GitHub. Read the source before you install it.

Installation

Install the Impact Trace 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/Prismer-AI/PrismerCloud/tree/main/sdk/apc/skills/impact-trace
  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/Prismer-AI/PrismerCloud.git /tmp/PrismerCloud
mkdir -p .claude/skills
cp -r /tmp/PrismerCloud/sdk/apc/skills/impact-trace .claude/skills/impact-trace
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Impact Trace 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 Impact Trace 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 Impact Trace 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.

impact-trace

给定一个改动点(函数 / 类型 / endpoint / SQL 列),产出它的真实影响面(受影响的调用方 / 契约消费者 / 持久化写入 / 双 Prisma schema / daemon-SDK 镜像 / skill-role 引用)+ 由此推导的回归测试范围(apc/05 §2 S2 · apc/12 impact-trace 行)。

什么时候用:一个 bugfix / feature 改完,进 review 前需要知道"改这一处会波及哪些地方、要回归哪些层";或评审要判断改动的 blast radius。

铁律:影响表的每一行都必须有 grep/search 证据(path+line)。凭记忆或推理写"这里应该会用到"是开卷考试——本 skill 的产出只认真实搜索命中。你自己不是 ground truth,rg 的命中集才是。

工具契约(签名以此为准,先核后用)

命令作用备注
rg <pattern> [path] -n主力搜索:直接调用方 / 契约消费者 / 引用点-n 出行号(证据必须带行号);-l 只列文件;-t ts 限类型
rg -c <pattern>命中计数(核对"没漏"用)计数对不上就是漏搜
cloud code grep <pattern> --repo <abs> [--glob '*.ts']白名单 repo 内搜索,输出 JSON [{path,line,snippet,sha256}]仅 PRISMER_WORKSPACE_ROOT 或 ~/.prismer/agents/<id>/scratch 内;出带 sha 的可引用片段。--glob 只支持 *.ext 形态
  • rg 是本机主力;cloud code grep 用于要产出可引用证据片段(带 sha256)时。两者的命中集应当一致——不一致就查是不是路径白名单/glob 把一部分挡在外面了。
  • 不要用单文件 grep 命中当全貌(memory feedback_grep_full_range_for_seq_audit)。搜完要核计数:rg -c 的总数 == 影响表行数(去掉 N/A 维度),对不上就是漏了。

六维影响面(逐维搜,缺一维要么有证据要么写 N/A 理由)

改动点波及面按六维穷举。任一维标 N/A 必须写理由;双 Prisma 维度永远不许静默省略(apc/12 负控:缺双 Prisma 维且无 N/A 理由 → 校验拒绝)。

#维度怎么搜(示例)找什么
1直接调用方rg '\b<name>\b' -n -t ts谁 import / 调用了这个符号
2契约消费者(API/wire)rg '<endpoint-path>|<event-code>|<rpc-name>' -nHTTP 路由消费者、SDK 类型、wire 帧的两端
3持久化写入rg '<column>|<model>\.(create|update|upsert)' -n谁写这张表/这个字段
4双 Prisma schemarg '<field>' prisma/schema.prisma prisma/schema.mysql.prisma -nSQLite + MySQL 两份 schema 是否都要改(不许省)
5daemon-SDK 镜像rg '<name>' sdk/prismer-cloud/runtime/src -ncloud 侧改了,daemon/SDK 侧有没有对称的一份要跟着改
6skill / role 引用rg '<name>' sdk/prismer-cloud/built-in-skills sdk/apc/skills -nSKILL.md / role template / prompt 里有没有引用它

Procedure

1. 锚定改动点

明确改动点的精确标识(符号名 / endpoint 路径 / 列名)。一个改动可能有多个改动点——逐个搜,别只搜一个。

2. 逐维搜索,收集带行号的命中

bash
# 维度 1:直接调用方(示例,把 <name> 换成真符号)
rg '\b<name>\b' -n -t ts > /tmp/impact-callers.txt
rg -c '\b<name>\b' -t ts    # 核对命中计数

# 维度 4:双 Prisma(两份 schema 都要看)
rg '<field>' prisma/schema.prisma prisma/schema.mysql.prisma -n

对每一维,把命中记成影响表的一行:{ path, line, dimension, contract, risk }。没有命中的维度:确认真的没有(不是搜错 pattern),然后写 N/A + 理由。

3. 组装影响表(副作用产出 1)

path:line维度契约风险回归
src/foo.ts:42直接调用方传参 shape改签名会破T0

每行必须能回指一条 rg/cloud code grep 命中。行数(不含 N/A 维)应等于去重后的真实命中数——自查:影响表行数 == Σ rg -c。

4. 推导回归测试范围(副作用产出 2)

按影响表命中的文件所在层映射到 tier(对齐 test-runner 选层规则):

命中落在回归 tier
src/lib/**T0
src/im/** / endpointT1
src/app/**T2
e2e / 跨端T3
daemon runtime/src/**daemon 测试(runtime vitest)

产出:"本改动需回归 {tiers},因为影响面落在 {层}"。

5.(可选)上报 criterion

若本 impact-trace 挂在一个 task 的 acceptance criterion 上,把结论上报:

bash
cloud task verify-criterion "$PRISMER_TASK_ID" "<criterion-id>" --outcome passed \
  --note "impact table: N rows, all grep-cited; regression scope: T0,T1"

--outcome 合法值恰好四个:passed | failed | n/a | waived。

输出契约(机器判据按这个复验,别自由发挥格式)

本 skill 的验收判据不是「报告里出现了某些词」,而是把报告里的声明拿回文件系统复核(structured-criteria.ts 的 cited-evidence / dimension-coverage)。所以报告必须带下面三样可机器解析的东西——形状对但内容假的报告会被当场判红。

① 改动点声明(一行,锚点必须是本次真实改动点,且会被复核确实存在于该文件):

CHANGE-POINT: <symbol> @ <path>

② 影响表(markdown 表,首列是 path:line):判据会逐行把 path:line 读回磁盘——文件不存在 / 行号越界 / 该行是空行,任意一条就判红;并且至少 3 条引用所在的那一行必须真的含改动点符号("真实存在的行" ≠ "真实的搜索命中"),且这些命中要落在 ≥3 个不同文件里——把一条真命中抄满整张表是"单命中报告",不是影响面(实证:一条真 path:line 复制 12 遍,加固前 5/5 全绿)。

| path:line | 维度 | 契约 | 风险 | 回归 |
| --- | --- | --- | --- | --- |
| src/foo.ts:42 | direct-callers | 传参 shape | 改签名会破 | T0 |

③ 六维逐行作答(每维独占一行 + 固定 key)。判据找的是这一行,报告别处出现同一个词(比如正文里出现 npm run prisma:generate)不算答过——这正是旧判据 match:"prisma" 被证伪的地方:

- [1] direct-callers: <path:line>
- [2] contract-consumers: <path:line>
- [3] persistence-writers: <path:line> | N/A — <理由 ≥20 字符>
- [4] dual-prisma: <path:line> | N/A — <理由 ≥20 字符>
- [5] daemon-sdk-mirror: <path:line>
- [6] skill-role-references: <path:line> | N/A — <理由 ≥20 字符>

N/A 必须带理由(去空白后 ≥20 字符);只写 N/A 判红。行号照抄 rg -n 的真实输出,不许凭印象写。 每一维要有自己的证据:同一条 path:line 最多给 2 个维度当证据,第 3 个维度再引它就判红(N/A 维不受影响,它本来就没有引用)。

产出(副作用 oracle,报告里必须给)

  1. 影响表:每行 {path:line, 维度, 契约, 风险} + 找到它的那条搜索。行的真实性 = grep 命中,不是叙述。
  2. 六维覆盖:六维全答;N/A 维带理由(双 Prisma 维不许静默省)。
  3. 回归范围:由影响表的层映射推出的 tier 集。 4.(若有 task)criterion 上报行:cloud task verify-criterion 落库 outcome,可回读 acceptance view 确认。

不许:把没有搜索命中的行写进影响表;漏搜一个真实调用方(rg -c 计数不核);把双 Prisma 维度静默省掉;断言聊天文本而非 path:line 证据。

诚实边界

  • rg 命中集是"文本匹配"的 ground truth,不是"语义调用图"——动态派发 / 字符串拼出来的引用 / 反射,文本搜不到。影响表覆盖的是可被文本搜索命中的引用面;无法静态搜到的动态引用要显式标注为已知盲区,不得声称"影响面已穷举"。
  • 回归范围是"应当回归哪些层"的推导,不代替真跑——真跑归 test-runner skill。

调用回执(APC 平台方运营纪律 · 自动注入,勿手改)

执行本 skill 的第一步,先落一条调用回执:

bash
cloud skill ack impact-trace --task "$PRISMER_TASK_ID"

按退出码分流(这条命令的退出码是承重信息,禁止用 || 兜底、; true、 set +e 或重定向把它抹掉):

exit含义你要做的
0回执已落库(im_task_logs.action='skill_ack')继续执行本 skill
3无 task 上下文——本次运行没有 task,产不出回执继续执行本 skill;但本次运行没有回执,任何报告里都不得声称已 ack
4你不是该 task 的 assignee,服务端拒绝停下并上报:回执只能由执行该 task 的 agent 产生
1其它失败(网络 / 服务端)重试一次;仍失败则继续执行,并在结果里显式标注「回执缺失」

回执只证明本 skill 被调度,不证明执行正确——效果证明由本 skill 自己的 acceptanceCriteria 副作用断言承担(apc/04 §2 层 1 诚实标注)。

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 Impact Trace AI skill do?

For a change point (function / type / endpoint / SQL column), enumerate its real blast radius across six dimensions with a grep citation per row, then derive the regression test scope. Every impact-table row must trace to a real search hit — never a self-asserted claim.

Why use Impact Trace on TypingMind?

Because you install it once and use it with any model. Impact Trace 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 Impact Trace in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/Prismer-AI/PrismerCloud/tree/main/sdk/apc/skills/impact-trace. 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 Impact Trace?

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 Impact Trace?

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

Is the Impact Trace AI skill free?

Yes. It is published on GitHub by Prismer-AI under the MIT 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.

View all

Set up your own AI workspace now

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