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 agentmods add skills/adambien/airails/concept-annotatornpx skills add AdamBien/airails --skill concept-annotatorgit clone --depth 1 https://github.com/AdamBien/airailsWrote 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/adambien/airails/concept-annotator)<a href="https://agentmods.dev/skills/adambien/airails/concept-annotator"><img src="https://agentmods.dev/badge/skills/adambien/airails/concept-annotator.svg" alt="Measured on agentmods" 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 | $0.00133 | $0.00695 |
| Opus 5 | $0.00067 | $0.00347 |
| Sonnet 5 | $0.00027 | $0.00139 |
| Haiku 4.5 | $0.00013 | $0.00069 |
Grade A, and why
concept-annotator 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 4d 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 — 41 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Concept Annotator
Carry the captured domain knowledge into the code itself. Every legacy package becomes self-describing — which concepts live here, where the package is heading, what the naming evidence suggests — in the exact file where the future sbce spec will live. A developer opening the package sees the migration state without leaving the IDE; an annotated system cannot quietly pretend the migration is done.
Inputs
migration/GLOSSARY.md— canonical terms and## Decisions: ground truth.migration/CONCEPTS.md— evidence pointers mapping concepts to packages and classes.migration/CARVING.md— when present, supplies target BC and moves per package.
Prefer running after /concept-clarifier; if glossary entries are still hypothesis-marked, carry the marker into the annotation verbatim — never launder a hypothesis into a fact.
Workflow
- List the Java packages of the source tree.
- Map concepts to packages via the CONCEPTS.md evidence pointers; note the local aliases per package.
- Per package, write
package-info.java— or update only the owned section of an existing one — from references/package-info-template.md. - Verify the build stays green: annotations are docs, never behavior.
Annotation Content
Per package, inside the owned section:
- Concepts — canonical term plus the aliases as they appear locally (
Contract — here: Cntrct, VTRG). - Candidate BC — single lowercase glossary token, or
unassigned. WithCARVING.mdpresent: Target BC plus the move rows affecting this package. - Refactoring hints — evidence-backed suggestions: mixed concepts → split candidate, names to rename to canonical terms, apparent boundary/control/entity leaning per class.
- See — pointers to the
migration/artifacts and open Q-ids touching this package.
Rules
- Docs only — never touch code, imports, or resources. The smallest possible change to a runnable system is no change.
- Own only the marked section (
## Migration Notes (concept-annotator)). Preserve existing package-info content and hand edits outside the marker; re-runs regenerate only the owned section. - Open every generated section with the disclaimer
Migration notes — not an sbce spec.— sbce treats package docs as authoritative boundary contracts, and these hints must never be mistaken for one./sbcelater replaces the notes with a real spec. - Hints are candidates, not verdicts: phrase as evidence-backed suggestions, never contradict the glossary, carry hypothesis markers and Q-ids instead of resolving them.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 4d ago First seen · 41 lines · 133 tokens per session scan A c7b3d6973341
concept-annotator is a skill published in the GitHub repository AdamBien/airails (48 stars, last pushed 14d ago), licensed MIT. It adds 133 tokens to every session and 695 once invoked, about $0.0007 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 2026-08-30.
Other skills, from other repositories
java-boilerplate
Gera automaticamente getters, setters, construtores, toString, equals e hashCode para classes Java. Triggers on: /java-boilerplate, 'gerar métodos java', 'criar getters setters'.
jeecg-dev
JeecgBoot 开发规范（仅手动触发）。⚠️ 本技能只在用户显式输入 /jeecg-dev 命令时使用，禁止自动触发——编写/修改 JeecgBoot 代码、应用 GitHub PR/issue 改动、修复 bug、新增功能、重构、代码生成等场景都不要自动调用本技能。内容涵盖 update-begin/end 痕迹注释、命名规范、实体/控制器/服务模式、API 约定、建表规则与修改日志实践。MANUAL ONLY: invoke ONLY when the user explicitly runs the /jeecg-dev command. Do NOT auto-trigger on any code editing…
deno-sandbox
Use when building features that execute untrusted user code, AI-generated code, or need isolated code execution environments. Covers the @deno/sandbox SDK.
deno-frontend
Use when building a web frontend with Deno — running React, Vite, Astro, SvelteKit, Next.js, Nuxt or other npm frameworks under Deno, or working with Fresh, Deno's own island-architecture framework. Covers which path to pick, Fresh 2.x routes, handlers, islands, Preact signals, Tailwind, and Fresh 1.x to 2.x migration.
5writing
数学建模竞赛论文撰写阶段，支持 Typst 和 LaTeX 双引擎。根据 ANALYSISMODELINGREPORT.md、RESULTSREPORT.md 和 figures/.pdf 选择比赛模板、排版引擎、组织章节，并在论文正文中按章节直接插入图表。.
6verity
数学建模竞赛最终验证和验收阶段，支持 Typst 和 LaTeX 双引擎。用于论文写完后检查章节数量、标题顺序、图表引用、数值一致性、占位符、内部文件泄露、参考文献、代码可复现性、编译和提交就绪状态。.