Add tempguru event staffing tool plugin by kissmyabs32 · Pull Request #3252 · langgenius/dify-official-plugins

kissmyabs32 · 2026-06-04T14:49:56Z

Summary

Adds the tempguru tool plugin under tools/tempguru/. Wraps the public TempGuru MCP REST API as 5 read-only Dify tools so any Dify agent can answer event-staffing questions (coverage, rates, lead times, state compliance) for events across 300+ US and Canadian markets — no authentication, no API key, no setup.

Matches the no-auth precedent set by duckduckgo and wikipedia in the same plugins directory.

What it does

Tool	Returns
`get_cities`	Cities TempGuru staffs, with tier classification. Filter by state or tier.
`get_roles`	Event staffing roles with descriptions and skill tiers.
`check_availability`	Lead-time guidance for a city + date. Not real-time inventory.
`get_role_pricing`	All-inclusive hourly rate range for a role in a city (W-2 wages, workers comp, payroll taxes, general liability all included).
`get_compliance_by_state`	US state-level minimum wage, overtime rules, compliance summary. Not legal advice.

Each tool is a ~20-line Python class that makes one HTTPS GET to mcp.tempguru.co/api/v1/<endpoint> and yields the response JSON back to Dify. No business logic reimplemented; the TempGuru MCP REST mirror is the canonical source.

Trust profile

Read-only — every tool wraps a GET. No POST / PUT / DELETE exists in the plugin or the underlying API.
No authentication — public data only. No API keys, OAuth, credentials.
No data collection — the plugin does not persist user inputs, queries, or outputs.
No executable scripts — only the standard Dify plugin entrypoint.
Same data as the public site — every response comes from TempGuru's published rate cards, coverage maps, and compliance summaries at https://tempguru.co.

Bilingual labels

All label, human_description, and description.human fields ship in both en_US and zh_Hans (Simplified Chinese), matching the convention of the highest-quality existing plugins. TempGuru already serves a Chinese-facing agent page at https://tempguru.co/zh-cn/ai.

Already shipped on adjacent surfaces

Official MCP Registry: co.tempguru/event-staffing (DNS-verified namespace)
Smithery: tempguru/event-staffing (quality score 90/100)
ModelScope MCP Plaza (Alibaba ecosystem): tempguru/TempGuru-Event-Staffing
Glama: connector listing live and Healthy
ClawHub, agentskill.sh, LobeHub
mcp.so issue [openrouter] Update model validation to use 'openrouter/free' #2625, mcpservers.org listing live
openai/skills upstream PR Feat add displayname for openai api compatible #472 (curated skill submission)

Documentation

Developer page: https://tempguru.co/ai (English) · https://tempguru.co/zh-cn/ai (Chinese)
MCP server source: https://github.com/kissmyabs32/tempguru-mcp
MCP server card: https://mcp.tempguru.co/.well-known/mcp/server-card.json
OpenAPI 3.1 spec: https://mcp.tempguru.co/openapi.json

License

MIT

Wraps the public TempGuru MCP REST API as 5 read-only Dify tools: - get_cities : list cities by state/tier - get_roles : list event staffing roles - check_availability : lead-time guidance for a city + date - get_role_pricing : all-inclusive hourly rate range - get_compliance_by_state : state-level wage/overtime/compliance Architecture: each tool is a 20-line Python class that makes one HTTPS GET to mcp.tempguru.co/api/v1/<endpoint> and yields the response JSON back to Dify. No business logic reimplemented; the MCP REST mirror is the canonical source. Trust profile: - Read-only (every tool is a GET) - No authentication required (public data) - No API keys, OAuth, credentials - No user data persisted - No executable scripts beyond the standard Dify entrypoint Matches the no-auth precedent set by duckduckgo and wikipedia plugins in the same .curated directory. Bilingual labels (en_US + zh_Hans) throughout.

gemini-code-assist

Code Review

This pull request introduces the TempGuru Event Staffing Dify plugin, which provides read-only access to event staffing data across US and Canadian markets through five tools. The feedback focuses on improving the robustness and safety of the plugin. Specifically, it is recommended to simplify credential validation by using a direct HTTP request rather than invoking internal SDK methods, and to clean up the resulting unused import. Additionally, the reviewer suggests implementing defensive parameter checks and wrapping API requests in try-except blocks across all tools to gracefully handle network errors and JSON parsing failures.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

gemini-code-assist · 2026-06-04T14:51:57Z

+    def _validate_credentials(self, credentials: dict) -> None:
+        try:
+            for _ in GetCitiesTool.from_credentials(credentials).invoke_from_executor(
+                tool_parameters={}
+            ):
+                pass
+        except Exception as e:
+            raise ToolProviderCredentialValidationError(
+                f"Unable to reach TempGuru MCP REST API at mcp.tempguru.co: {str(e)}"
+            )


Instantiating a tool class and calling internal SDK methods like invoke_from_executor inside _validate_credentials is highly risky and prone to breaking if the SDK internals change. It is much safer and cleaner to perform a direct HTTP request to the public API endpoint to verify connectivity.

Suggested change

def _validate_credentials(self, credentials: dict) -> None:

try:

for _ in GetCitiesTool.from_credentials(credentials).invoke_from_executor(

tool_parameters={}

):

pass

except Exception as e:

raise ToolProviderCredentialValidationError(

f"Unable to reach TempGuru MCP REST API at mcp.tempguru.co: {str(e)}"

)

def _validate_credentials(self, credentials: dict) -> None:

try:

response = requests.get(

"https://mcp.tempguru.co/api/v1/cities",

headers={"User-Agent": "tempguru-dify-plugin/0.0.1", "Accept": "application/json"},

timeout=10,

)

response.raise_for_status()

except Exception as e:

raise ToolProviderCredentialValidationError(

f"Unable to reach TempGuru MCP REST API at mcp.tempguru.co: {str(e)}"

)

gemini-code-assist · 2026-06-04T14:51:57Z

+from dify_plugin.errors.tool import ToolProviderCredentialValidationError
+from dify_plugin import ToolProvider
+from tools.get_cities import GetCitiesTool


The import of GetCitiesTool is unused if we simplify the credential validation logic to use a direct HTTP request. Removing it keeps the imports clean and avoids potential import cycle or path resolution issues.

Suggested change

from dify_plugin.errors.tool import ToolProviderCredentialValidationError

from dify_plugin import ToolProvider

from tools.get_cities import GetCitiesTool

from dify_plugin.errors.tool import ToolProviderCredentialValidationError

from dify_plugin import ToolProvider

import requests

gemini-code-assist · 2026-06-04T14:51:57Z

+    def _invoke(
+        self, tool_parameters: dict[str, Any]
+    ) -> Generator[ToolInvokeMessage, None, None]:
+        params: dict[str, Any] = {
+            "city": tool_parameters["city"],
+            "date": tool_parameters["date"],
+        }
+        if tool_parameters.get("role"):
+            params["role"] = tool_parameters["role"]
+        if tool_parameters.get("count"):
+            params["count"] = tool_parameters["count"]
+
+        response = requests.get(
+            f"{API_BASE}/availability",
+            params=params,
+            headers={"User-Agent": USER_AGENT, "Accept": "application/json"},
+            timeout=30,
+        )
+        response.raise_for_status()
+        yield self.create_json_message(response.json())


To ensure robust error handling and defensive programming, we should safely extract parameters using .get() and check for their presence. Additionally, wrapping the HTTP request in a try-except block to catch requests.RequestException and ValueError prevents unhandled exceptions from crashing the tool execution, allowing us to return a clean error message instead.

def _invoke( self, tool_parameters: dict[str, Any] ) -> Generator[ToolInvokeMessage, None, None]: city = tool_parameters.get("city") date = tool_parameters.get("date") if not city or not date: yield self.create_text_message("Both 'city' and 'date' parameters are required.") return params: dict[str, Any] = { "city": city, "date": date, } if tool_parameters.get("role"): params["role"] = tool_parameters["role"] if tool_parameters.get("count"): params["count"] = tool_parameters["count"] try: response = requests.get( f"{API_BASE}/availability", params=params, headers={"User-Agent": USER_AGENT, "Accept": "application/json"}, timeout=30, ) response.raise_for_status() yield self.create_json_message(response.json()) except requests.RequestException as e: yield self.create_text_message(f"API request failed: {str(e)}") except ValueError: yield self.create_text_message("Failed to parse response from TempGuru API.")

gemini-code-assist · 2026-06-04T14:51:57Z

+    def _invoke(
+        self, tool_parameters: dict[str, Any]
+    ) -> Generator[ToolInvokeMessage, None, None]:
+        params: dict[str, Any] = {}
+        if tool_parameters.get("state"):
+            params["state"] = tool_parameters["state"]
+        if tool_parameters.get("tier"):
+            params["tier"] = tool_parameters["tier"]
+
+        response = requests.get(
+            f"{API_BASE}/cities",
+            params=params,
+            headers={"User-Agent": USER_AGENT, "Accept": "application/json"},
+            timeout=30,
+        )
+        response.raise_for_status()
+        yield self.create_json_message(response.json())


Wrapping the network request in a try-except block ensures that any connection issues, timeouts, or invalid JSON responses are handled gracefully rather than raising unhandled exceptions.

def _invoke( self, tool_parameters: dict[str, Any] ) -> Generator[ToolInvokeMessage, None, None]: params: dict[str, Any] = {} if tool_parameters.get("state"): params["state"] = tool_parameters["state"] if tool_parameters.get("tier"): params["tier"] = tool_parameters["tier"] try: response = requests.get( f"{API_BASE}/cities", params=params, headers={"User-Agent": USER_AGENT, "Accept": "application/json"}, timeout=30, ) response.raise_for_status() yield self.create_json_message(response.json()) except requests.RequestException as e: yield self.create_text_message(f"API request failed: {str(e)}") except ValueError: yield self.create_text_message("Failed to parse response from TempGuru API.")

gemini-code-assist · 2026-06-04T14:51:57Z

+    def _invoke(
+        self, tool_parameters: dict[str, Any]
+    ) -> Generator[ToolInvokeMessage, None, None]:
+        params = {"state": tool_parameters["state"]}
+        response = requests.get(
+            f"{API_BASE}/compliance",
+            params=params,
+            headers={"User-Agent": USER_AGENT, "Accept": "application/json"},
+            timeout=30,
+        )
+        response.raise_for_status()
+        yield self.create_json_message(response.json())


Safely retrieve the required 'state' parameter and wrap the HTTP request in a try-except block to handle potential network errors or parsing failures gracefully.

def _invoke( self, tool_parameters: dict[str, Any] ) -> Generator[ToolInvokeMessage, None, None]: state = tool_parameters.get("state") if not state: yield self.create_text_message("The 'state' parameter is required.") return params = {"state": state} try: response = requests.get( f"{API_BASE}/compliance", params=params, headers={"User-Agent": USER_AGENT, "Accept": "application/json"}, timeout=30, ) response.raise_for_status() yield self.create_json_message(response.json()) except requests.RequestException as e: yield self.create_text_message(f"API request failed: {str(e)}") except ValueError: yield self.create_text_message("Failed to parse response from TempGuru API.")

gemini-code-assist · 2026-06-04T14:51:57Z

+    def _invoke(
+        self, tool_parameters: dict[str, Any]
+    ) -> Generator[ToolInvokeMessage, None, None]:
+        params = {
+            "role": tool_parameters["role"],
+            "city": tool_parameters["city"],
+        }
+        response = requests.get(
+            f"{API_BASE}/pricing",
+            params=params,
+            headers={"User-Agent": USER_AGENT, "Accept": "application/json"},
+            timeout=30,
+        )
+        response.raise_for_status()
+        yield self.create_json_message(response.json())


Perform defensive parameter checks and wrap the HTTP request in a try-except block to handle network errors or invalid JSON responses gracefully.

def _invoke( self, tool_parameters: dict[str, Any] ) -> Generator[ToolInvokeMessage, None, None]: role = tool_parameters.get("role") city = tool_parameters.get("city") if not role or not city: yield self.create_text_message("Both 'role' and 'city' parameters are required.") return params = { "role": role, "city": city, } try: response = requests.get( f"{API_BASE}/pricing", params=params, headers={"User-Agent": USER_AGENT, "Accept": "application/json"}, timeout=30, ) response.raise_for_status() yield self.create_json_message(response.json()) except requests.RequestException as e: yield self.create_text_message(f"API request failed: {str(e)}") except ValueError: yield self.create_text_message("Failed to parse response from TempGuru API.")

gemini-code-assist · 2026-06-04T14:51:57Z

+    def _invoke(
+        self, tool_parameters: dict[str, Any]
+    ) -> Generator[ToolInvokeMessage, None, None]:
+        response = requests.get(
+            f"{API_BASE}/roles",
+            headers={"User-Agent": USER_AGENT, "Accept": "application/json"},
+            timeout=30,
+        )
+        response.raise_for_status()
+        yield self.create_json_message(response.json())


Wrap the HTTP request in a try-except block to handle network errors or invalid JSON responses gracefully.

Suggested change

def _invoke(

self, tool_parameters: dict[str, Any]

) -> Generator[ToolInvokeMessage, None, None]:

response = requests.get(

f"{API_BASE}/roles",

headers={"User-Agent": USER_AGENT, "Accept": "application/json"},

timeout=30,

)

response.raise_for_status()

yield self.create_json_message(response.json())

def _invoke(

self, tool_parameters: dict[str, Any]

) -> Generator[ToolInvokeMessage, None, None]:

try:

response = requests.get(

f"{API_BASE}/roles",

headers={"User-Agent": USER_AGENT, "Accept": "application/json"},

timeout=30,

)

response.raise_for_status()

yield self.create_json_message(response.json())

except requests.RequestException as e:

yield self.create_text_message(f"API request failed: {str(e)}")

except ValueError:

yield self.create_text_message("Failed to parse response from TempGuru API.")

…vider validation Applies the 7 suggestions from the Gemini Code Assist bot on PR langgenius#3252: provider/tempguru.py - Replace SDK-internal-method validation with a direct requests.get probe. Avoids coupling the validator to Tool subclass internals. - Remove unused GetCitiesTool import; add requests. All 5 tool handlers - Wrap requests.get in try/except for requests.RequestException and ValueError. Yield create_text_message on errors instead of raising. Tools with required parameters - Use tool_parameters.get() instead of subscript access on required params. Return early with a helpful error message if any required param is missing.

kissmyabs32 · 2026-06-04T14:55:47Z

Thanks for the thorough review. All 7 suggestions applied in 17a231d4:

provider/tempguru.py:

Switched from GetCitiesTool.from_credentials().invoke_from_executor() to a direct requests.get probe — agreed that internal-SDK coupling is the wrong dependency to take
Removed the now-unused GetCitiesTool import
Added import requests

All 5 tool handlers:

Wrapped requests.get in try/except for requests.RequestException and ValueError
Yield self.create_text_message(...) on errors so the agent sees a clean message instead of an unhandled exception

Tools with required params (check_availability, get_role_pricing, get_compliance_by_state):

Now use tool_parameters.get() with a None check + early-return helpful error message, rather than subscript access that would KeyError if a param is missing

Ready for re-review.

CI's astral-sh/setup-uv@v6 looks for uv.lock or requirements.txt to invalidate the dependency cache. Without either, the cache key check fails with 'No file matched to .../uv.lock,.../requirements.txt' and the downstream test step exits 1. Generated by 'uv lock' against pyproject.toml. Resolved 41 packages including dify_plugin transitive deps. Lockfile is portable across Python 3.12+ per requires-python in pyproject.toml.

The pre-pr-check-per-plugin workflow at .github/workflows/ line 120 hard-requires the top-level manifest.yaml author field to match 'langgenius' (the org maintaining the official-plugins monorepo). Individual contributor attribution stays in provider/<name>.yaml which retains 'author: tempguru' — same pattern as duckduckgo (provider author: yash_parmar) and other community contributions.

dosubot Bot added size:L This PR changes 100-499 lines, ignoring generated files. enhancement New feature or request labels Jun 4, 2026

kissmyabs32 had a problem deploying to tools/tempguru June 4, 2026 14:50 — with GitHub Actions Failure

gemini-code-assist Bot reviewed Jun 4, 2026

View reviewed changes

kissmyabs32 had a problem deploying to tools/tempguru June 4, 2026 14:56 — with GitHub Actions Failure

kissmyabs32 had a problem deploying to tools/tempguru June 4, 2026 14:58 — with GitHub Actions Failure

kissmyabs32 deployed to tools/tempguru June 4, 2026 15:01 — with GitHub Actions Active

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add tempguru event staffing tool plugin#3252

Add tempguru event staffing tool plugin#3252
kissmyabs32 wants to merge 4 commits into
langgenius:mainfrom
kissmyabs32:add-tempguru-event-staffing

kissmyabs32 commented Jun 4, 2026

Uh oh!

gemini-code-assist Bot left a comment

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Uh oh!

kissmyabs32 commented Jun 4, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

kissmyabs32 commented Jun 4, 2026

Summary

What it does

Trust profile

Bilingual labels

Already shipped on adjacent surfaces

Documentation

License

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jun 4, 2026

Choose a reason for hiding this comment

Uh oh!

kissmyabs32 commented Jun 4, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant