model updatae skill#414
Merged
c121914yu merged 2 commits intolabring:mainfrom May 6, 2026
Merged
Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 46ade588cc
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| list: [ | ||
| { | ||
| type: ModelTypeEnum.llm, | ||
| model: 'gpt-5.5', |
There was a problem hiding this comment.
When a user selects this new preset, FastGPT will send gpt-5.5 (and the adjacent gpt-5.5-pro) to the OpenAI API, but OpenAI's current model catalog lists GPT-5.2/GPT-5.1/GPT-5 and GPT-5.2-pro/GPT-5-pro rather than these IDs (checked https://platform.openai.com/docs/models). These presets will therefore fail at runtime for anyone choosing them.
Useful? React with 👍 / 👎.
| { | ||
| type: ModelTypeEnum.llm, | ||
| model: 'gemini-3-pro-preview', | ||
| model: 'gemini-3-flash', |
There was a problem hiding this comment.
For Gemini users who need the Pro preview, this edit removes the valid gemini-3-pro-preview preset by changing that slot to gemini-3-flash; Google's Gemini API docs list the API IDs as gemini-3-pro-preview, gemini-3-flash-preview, and gemini-3-pro-image-preview for the Gemini 3 series (checked https://ai.google.dev/gemini-api/docs/models/gemini). As a result the Pro model disappears from FastGPT, while the newly exposed non-preview ID may be rejected by the API.
Useful? React with 👍 / 👎.
|
✅ Build Successful - Preview Images for this PR: Changed packages: |
Contributor
There was a problem hiding this comment.
Pull request overview
This PR updates the FastGPT plugin’s model provider preset registry (new/updated model IDs and rerank metadata), bumps the SDK package version, and introduces new .codex skill documentation/scripts while removing the older .claude skill assets.
Changes:
- Add/refresh multiple provider presets (OpenAI/Claude/Gemini/Groq/Grok/MistralAI/StepFun, plus rerank/STT tweaks) and remove some deprecated Moonshot “kimi-latest-*” entries.
- Extend rerank model validation to require
maxToken, and backfillmaxTokenon existing rerank presets. - Add
AGENTS.mdplus new.codex/skills/*skill docs and helper scripts; remove.claude/skills/toolDev/*.
Reviewed changes
Copilot reviewed 31 out of 33 changed files in this pull request and generated 4 comments.
Show a summary per file
| File | Description |
|---|---|
| sdk/package.json | Bump SDK version to 0.6.1. |
| modules/model/provider/StepFun/index.ts | Add new StepFun LLM presets. |
| modules/model/provider/Siliconflow/index.ts | Add maxToken to rerank preset. |
| modules/model/provider/Qwen/index.ts | Add new rerank preset and add maxToken to existing rerank. |
| modules/model/provider/OpenAI/index.ts | Add new OpenAI LLM presets and small formatting cleanup. |
| modules/model/provider/Moonshot/index.ts | Remove kimi-latest-* LLM presets. |
| modules/model/provider/MistralAI/index.ts | Add new MistralAI LLM presets. |
| modules/model/provider/Jina/index.ts | Add maxToken to rerank presets. |
| modules/model/provider/Groq/index.ts | Add new Groq LLM and STT presets. |
| modules/model/provider/Grok/index.ts | Add new Grok LLM preset. |
| modules/model/provider/Gemini/index.ts | Reorder/rename Gemini model presets. |
| modules/model/provider/Claude/index.ts | Add new Claude LLM presets. |
| modules/model/provider/BAAI/index.ts | Add maxToken to rerank preset. |
| modules/model/provider/AliCloud/index.ts | Add fun-asr STT preset. |
| lib/validates/model.ts | Require maxToken in rerank schema. |
| AGENTS.md | Add contributor/development guide. |
| .codex/skills/tool-dev/SKILL.md | Add “tool-dev” skill documentation. |
| .codex/skills/tool-dev/scripts/validate_tool_package.mjs | Add tool package structural validation script. |
| .codex/skills/tool-dev/references/requirement_template.md | Add tool design template. |
| .codex/skills/tool-dev/references/design_spec.md | Add tool design reference/spec. |
| .codex/skills/tool-dev/agents/openai.yaml | Add agent metadata for “tool-dev” skill. |
| .codex/skills/model-provider-updater/SKILL.md | Add “model-provider-updater” skill documentation. |
| .codex/skills/model-provider-updater/scripts/model_provider_presets.mjs | Add provider preset inventory/plan/apply helper script. |
| .codex/skills/model-provider-updater/references/provider_sources.json | Add official provider source hints. |
| .codex/skills/model-provider-updater/references/plan.example.json | Add example plan format. |
| .codex/skills/model-provider-updater/agents/openai.yaml | Add agent metadata for “model-provider-updater” skill. |
| .claude/skills/toolDev/SKILL.md | Removed legacy skill documentation. |
| .claude/skills/toolDev/scripts/quick_validate.py | Removed legacy validator script. |
| .claude/skills/toolDev/scripts/package_skill.py | Removed legacy packager script. |
| .claude/skills/toolDev/scripts/init_skill.py | Removed legacy initializer script. |
| .claude/skills/toolDev/references/requirement_template.md | Removed legacy requirement template. |
| .claude/skills/toolDev/references/design_spec.md | Removed legacy design spec. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| Run it before final validation for new or heavily edited tool packages: | ||
|
|
||
| ```bash | ||
| node .agents/skills/tool-dev/scripts/validate_tool_package.mjs modules/tool/packages/<toolName> |
|
|
||
| ## Validation | ||
|
|
||
| - `node .agents/skills/tool-dev/scripts/validate_tool_package.mjs modules/tool/packages/<toolName>` |
Comment on lines
+13
to
+46
| 1. Inventory the current registry. | ||
|
|
||
| ```bash | ||
| node .agents/skills/model-provider-updater/scripts/model_provider_presets.mjs inventory | ||
| ``` | ||
|
|
||
| Use `--json` when another script or a temporary comparison file needs structured output. | ||
|
|
||
| 2. Check providers one by one against official sources. | ||
|
|
||
| Start from `references/provider_sources.json`, but verify the current official page/API during the run because model catalogs change often. Use official provider docs, official pricing/model pages, or official model-list APIs. Do not use third-party blogs, search snippets, or aggregator pages as removal evidence unless the provider is that aggregator, such as OpenRouter. | ||
|
|
||
| `check-sources` only checks whether source-hint URLs are reachable enough to use as starting points. A 403 access-limited result can still be acceptable for docs that block automated requests, and a passing source check is not evidence that the provider catalog was audited. | ||
|
|
||
| 3. Classify differences conservatively. | ||
|
|
||
| Add a preset when an official source lists a model that is absent locally and it belongs to an existing provider. Clone the closest existing preset in the same provider and family, then adjust context, output limit, vision, reasoning, tool calling, response-format, and field-map fields from official docs or the closest local pattern. | ||
|
|
||
| Remove a preset only when an official source explicitly marks the model as deprecated, retired, unavailable, or when an authoritative model-list API/document states that only the returned/listed models are supported and the local model is absent. Do not delete local custom placeholders in `Other`, `Ollama`, `HuggingFace`, `OpenRouter`, or similar open catalogs unless the provider explicitly removes that exact model from its own official API/catalog. | ||
|
|
||
| 4. Create and apply an update plan. | ||
|
|
||
| Generate a template and fill only confirmed additions/removals: | ||
|
|
||
| ```bash | ||
| node .agents/skills/model-provider-updater/scripts/model_provider_presets.mjs plan-template --provider OpenAI > /tmp/model-provider-plan.json | ||
| ``` | ||
|
|
||
| Dry-run before writing: | ||
|
|
||
| ```bash | ||
| node .agents/skills/model-provider-updater/scripts/model_provider_presets.mjs apply-plan --plan /tmp/model-provider-plan.json --dry-run | ||
| node .agents/skills/model-provider-updater/scripts/model_provider_presets.mjs apply-plan --plan /tmp/model-provider-plan.json --write | ||
| ``` |
Comment on lines
+845
to
+851
| function printUsage() { | ||
| console.log(`Usage: | ||
| node .agents/skills/model-provider-updater/scripts/model_provider_presets.mjs inventory [--provider OpenAI] [--json] | ||
| node .agents/skills/model-provider-updater/scripts/model_provider_presets.mjs plan-template [--provider OpenAI] | ||
| node .agents/skills/model-provider-updater/scripts/model_provider_presets.mjs apply-plan --plan /tmp/plan.json (--dry-run | --write) [--json] | ||
| node .agents/skills/model-provider-updater/scripts/model_provider_presets.mjs check-sources [--provider OpenAI] [--strict] [--json] [--concurrency 4] | ||
|
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.