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/benedictking/ccx/model-updatenpx skills add BenedictKing/ccx --skill model-updategit clone --depth 1 https://github.com/BenedictKing/ccxWhat 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 | $0.00043 | $0.01696 |
| Opus 5 | $0.00022 | $0.00848 |
| Sonnet 5 | $0.00009 | $0.00339 |
| Haiku 4.5 | $0.00004 | $0.00170 |
Grade A, and why
model-update 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 2d 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 — 173 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CCX 模型注册更新技能
当用户输入包含以下关键词时触发:
- "新模型"、"添加模型"、"model registry"、"模型注册"、"模型版本更新"、"model update"
适用范围
适用于两类场景:
- 新增模型族:仓库中还没有该 canonical model,需要补完整能力定义与必要映射。
- 已有模型族发布新 dated suffix / 版本别名:如
DeepSeek-V4-Pro-0813,通常只需要放宽 pattern、补测试,并检查特定 route 白名单是否也要同步放开。
执行流程
0. 先判定属于哪一类变更
先回答这 4 个问题:
- 仓库里是否已经存在 canonical model(如
deepseek-v4-pro)? - 新上线的是全新 canonical model,还是已有模型的新版本别名/日期后缀?
- 是否有 provider route 对该模型做了
SupportedModels白名单限制? - 是否有 benchmark / preset / builtin manifest 依赖生成产物自动同步?
如果只是 dated suffix 更新,优先最小化修改,避免误改优先级和 alias。
1. 收集模型信息
优先从官方文档、项目上下文或现有同族模型推断以下字段:
| 字段 | 示例 |
|---|---|
| canonical model | deepseek-v4-pro |
| 新上线别名 | DeepSeek-V4-Pro-0813 |
| provider | deepseek / anthropic / openai |
| 是否已有同族条目 | 是 / 否 |
| 是否只是日期后缀 | 是 / 否 |
| 是否影响 bare alias | 如 sonnet / opus / deepseek-chat |
| 是否影响特定协议路由白名单 | 如 Responses route |
2. 更新 single source of truth
核心文件: shared/model-registry/ccx_model_registry.json
2.1 如果是新增 canonical model
在 upstreamCapabilities 中新增完整条目,按家族就近放置。
2.2 如果是 dated suffix / 版本别名更新
优先修改已有 patterns,不要重复新增能力条目。
典型规则:
- 仅支持
-YYYY-MM-DD与-YYMMDD/-YYYYMMDD时,pattern 常是-\\d{6,8} - 若 DeepSeek 新版本使用
-MMDD,需放宽到-\\d{4,8}
例如:
"(?:^|[-/])deepseek-v4-pro(?:-\\d{4}-\\d{2}-\\d{2}|-\\d{4,8})?(?=$|@)"
3. 检查是否需要改 route 白名单
重点文件: backend-go/internal/config/provider_templates.go
检查:
SupportedModels- provider 描述文案
- route 描述文案
- 对应测试
backend-go/internal/config/provider_templates_test.go
典型场景:
- 上游某协议端点此前只支持
flash - 新版本上线后
pro也支持该端点 - 此时不仅要放宽注册表 pattern,还要把 route 白名单从
[]string{"deepseek-v4-flash"}扩成[]string{"deepseek-v4-flash", "deepseek-v4-pro"}
4. 仅在确有需要时更新其他源文件
按模型类型选择,不要机械全改:
Claude 系列常见附加点
backend-go/internal/config/model_registry.go里的 AgentModelProfileshared/model-priority/model-priority.ts- bare alias(如
sonnet/opus)
DeepSeek / 其他非 Claude 系列常见附加点
- benchmark 映射(若是新 canonical model)
- preset / builtin manifest(通常由生成脚本自动产出)
- provider template 的
SupportedModels
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.
- 2d ago First seen · 173 lines · 43 tokens per session scan A 75c4a1cde4bb
model-update is a skill published in the GitHub repository BenedictKing/ccx (3,966 stars, last pushed 2d ago), licensed MIT. It adds 43 tokens to every session and 1,696 once invoked, about $0.0002 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
spec_generator
Generates a structured Workable Spec JSON to guide a Developer Worker.
behavioral-evals
Guidance for creating, running, fixing, and promoting behavioral evaluations. Use when verifying agent decision logic, debugging failures, debugging prompt steering, or adding workspace regression tests.
google-agents-cli-observability
This skill should be used when the user wants to "set up tracing", "monitor my agent", "configure logging", "add observability", "debug production traffic", or needs guidance on monitoring deployed agents, including ADK (Agent Development Kit) agents. Covers Cloud Trace, prompt-response logging, BigQuery Agent…
google-agents-cli-publish
This skill should be used when the user wants to "publish an agent", "publish my ADK agent", "register an agent with Gemini Enterprise", "publish to Gemini Enterprise", or needs guidance on the agents-cli publish gemini-enterprise command. Also use when the user wants to "manage agents in Agent Registry"…
google-agents-cli-scaffold
This skill should be used when the user wants to "create an agent project", "start a new ADK project", "build me a new agent", "add CI/CD to my project", "add deployment", "enhance my project", or "upgrade my project". Part of the agents-cli skills suite. Covers agents-cli scaffold create, scaffold enhance, and…
agent-integration
Run all three agent integration phases sequentially: research, write-tests, and implement using E2E-first TDD (unit tests written last). For individual phases, use /agent-integration:research, /agent-integration:write-tests, or /agent-integration:implement. Use when the user says "integrate agent", "add agent…