Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add SummerSec/SumSec-Skills --skill lexicon-managergit clone --depth 1 https://github.com/SummerSec/SumSec-SkillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/summersec/sumsec-skills/lexicon-manager)<a href="https://agentmods.dev/skills/summersec/sumsec-skills/lexicon-manager"><img src="https://agentmods.dev/badge/skills/summersec/sumsec-skills/lexicon-manager/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/summersec/sumsec-skills/lexicon-manager"><img src="https://agentmods.dev/badge/skills/summersec/sumsec-skills/lexicon-manager.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00062 | $0.01457 |
| Opus 5 | $0.00031 | $0.00728 |
| Sonnet 5 | $0.00012 | $0.00291 |
| Haiku 4.5 | $0.00006 | $0.00146 |
Grade A, and why
lexicon-manager scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.
语义陷阱词典管理器
交互式维护词典:权威来源为 references/semantic-trap-lexicon.md 中的表格;lib/lexicon-data.js 由 npm run build-lexicon 自动生成,不要作为手写主路径(避免与校验脚本不一致)。
数据文件与生成命令
| 角色 | 路径 |
|---|---|
| 权威源(人编辑) | references/semantic-trap-lexicon.md |
| 生成物(勿手改为主流程) | lib/lexicon-data.js |
# 修改 MD 表格后执行:重新生成 lexicon-data.js
npm run build-lexicon
# 校验已提交的 lexicon-data.js 与 MD 一致(不写盘)
npm run build-lexicon:check
词汇对数据结构(写入 MD / 心智模型)
字段需能在表格中表达;生成脚本会解析为下列结构(与 lexicon-data.js 中对象一致)。
中文词汇对 (zhPairs)
id:T01–T99narrow/wide:窄边界 / 宽边界(中文)narrowEn/wideEn:括号内英文,与 MD 表格列一致severity:由 MD「语义宽度差」列映射为critical/high/medium-high/mediumscenario:失控场景(表格最后一列)
英文词汇对 (enPairs)
id:E01–E99narrow/wide;可选narrowAlt/wideAlt(MD 中用/分隔)severity:与仓库生成规则一致(由维护脚本映射)scenario
严重等级(severity)
critical > high > medium-high > medium > low
选择依据(摘录):
- critical:替换后输出量或范围变化极大,或明显偏离任务目标
- high:明显主观判断或超范围联想
- medium-high:适度发散仍在相关域内
- medium:边界略模糊、影响有限
- low:理论有差、实际影响极小
维护工作流
STEP 1:读取当前词典
- 读取
references/semantic-trap-lexicon.md中各表格,或require('./lib/lexicon-data.js')使用已生成的wideWordsZh/wideWordsEn向用户展示概览(二者应一致;若不一致先运行npm run build-lexicon:check定位)。
概览示例:
📖 当前词典概览
中文词汇对: T01–T17(以 MD 为准)
英文词汇对: E01–E10(以 MD 为准)
STEP 2:确认操作
与用户确认操作类型:添加 / 修改 / 删除。
操作 A:添加词汇对
逐项收集:
- 语言:中文 (T) 或英文 (E)
- wide / narrow(及中文的 narrowEn、wideEn;英文的 narrowAlt、wideAlt 若需要)
- severity(或对应 MD 语义宽度差文案)
- scenario(失控说明,非空)
ID:新 ID = 同系列当前最大 ID + 1(如最大 T17 → T18)。不随意重用已删除 ID(保持历史稳定)。
验证:
- wide / narrow 非空;wide 不与已有行冲突
- severity 合法;scenario 非空
操作 B:修改词汇对
- 用户指定 ID(如 T05)
- 展示当前表格行或生成条目中的字段
- 确认新值后更新 MD 表格对应行
操作 C:删除词汇对
- 用户指定 ID,展示完整信息
- 二次确认删除(降低检测覆盖)
- 从 MD 表格移除该行;不重新编号已有 ID
STEP 3:执行修改(仅改 MD → 生成)
- 编辑
references/semantic-trap-lexicon.md- 在「高危词汇对」「扩展词汇对」「英文环境高危词汇对」等对应表格中 增/改/删 行,列格式与现有行保持一致。
- 若变更影响说明,检查「LLM 语义敏感度矩阵」是否需同步调整。
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 138 lines · 62 tokens per session scan A b75e390523d6
lexicon-manager is a skill published in the GitHub repository SummerSec/SumSec-Skills (8 stars, last pushed 24d ago), licensed Apache-2.0. It adds 62 tokens to every session and 1,457 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
dsh-sdk-upgrade
Safely select and install a compatible official @deepseek-ai SDK release for dsh plugin projects (dsh-web, dsh-trading, and similar monorepos) from npm using an isolated worktree, explicit cohort review, CI-equivalent validation, and controlled rollout — including syncing the project's declared DSH host-version floor…
dsh-web-sdk-compatibility
Adapt and repair dsh-web after an approved official @deepseek-ai SDK/runtime cohort is selected or installed. Compare public API, type, service-injection, module-table, protocol, and behavior changes; map every change to repository consumers; implement the smallest fixes and durable compatibility contracts; handle…
ov-experience-memory
Retrieve and apply OpenViking Experience memories through the Agent runtime's generic OpenViking search and read tools. Use before or during executable, multi-step, or tool-based work such as coding, file or data changes, configuration, deployment, workflow execution, and failure recovery when prior operational…
audit-playbook
A code-security audit playbook covering source-code review, runtime checks, software dependencies, deployment settings, and AI-agent security risks.
x64dbg-reversing
Perform dynamic reverse engineering and debugging using X64dbg/X32dbg through natural language commands powered by X64dbg MCP. Use this skill whenever the user wants to: debug executables, analyze program behavior at runtime, set breakpoints, inspect memory and registers, trace execution flow, bypass anti-debug…
ctf-playbook
A playbook for solving capture-the-flag challenges, where competitors find a hidden proof string called a flag.