blackoutta/dify-skills-evaluator/dify-api-curl
Claude Code and agent skill by blackoutta/dify-skills-evaluator
Trust tier 3
- community
Source check pending
Checking that the SKILL.md file is still in the repository.
Build and debug Dify Service API requests with curl across completion, chat, advanced chat, and workflow apps. Use when Codex needs to choose the right Dify app API, inspect app parameters, stream SSE responses, upload files, manage conversations, stop running tasks, fetch workflow logs, or operate annotation endpoints while keeping curl as the primary caller.
Install
npx vskill@latest install blackoutta/dify-skills-evaluator/dify-api-curlSKILL.md
Dify API Curl
Use this skill to translate Dify Service API tasks into ready-to-run curl commands.
Workflow Decision Tree
- Read references/skills.md first.
- Read references/common.md for auth, environment variables, response-mode rules, state carry-forward, SSE, uploads, and shared endpoints.
- Choose one app family:
- Read references/completion.md for stateless text generation with
/completion-messages. - Read references/chat.md for session-based chat, conversation history, suggestions, and conversation variables.
- Read references/advanced_chat.md for chat requests that expose workflow/node trace events or annotation management.
- Read references/workflow.md for non-session workflow execution, specific workflow versions, run detail, and logs.
Operating Rules
- Prefer
curlover SDK snippets unless the user explicitly asks for another client. - Call
GET /parametersbefore composing a request if the required input variable names are unknown. - Respect the caller's required
response_mode. Do not silently switch a requiredblockingcall tostreaming. - Read timeout flags from environment variables instead of hardcoding them per command when possible.
- Use
curl -Nonly forresponse_mode=streamingso SSE frames are not buffered. - Preserve and surface returned identifiers:
task_id,message_id,conversation_id, andworkflow_run_id. - For chat-family follow-up turns, extract
conversation_idfrom the first response and reuse that exact value in the second request. Never leaveconversation_idempty when continuing a conversation. - Keep
userstable across turns of the same chat or workflow session. - Upload local files first with
/files/upload, then referenceupload_file_idin the follow-up request. - Use
X-Trace-Idwhen the user wants request tracing; otherwise body/querytrace_idis acceptable but lower priority.
Output Shape
When answering with this skill:
- Pick the correct app family explicitly.
- If
/inforeturnsmode: "advanced-chat", prefer references/advanced_chat.md even though the endpoint is still/chat-messages. - Show the minimal environment variables required to run the request.
- Return one primary
curlexample first. - Add follow-up
curlcommands only for the next likely step: stop, fetch history, list logs, rename conversation, or inspect parameters. - For multi-turn chat tasks, show the handoff step that stores
conversation_idbetween turns. - Mention the key streaming events or response fields that matter for that app family.
Security
- Security scanPassed, 95 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/dify-api-curl/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/blackoutta/dify-skills-evaluator/dify-api-curl)