Tendem MCP lets you delegate tasks to vetted human experts directly from your AI assistant, managing the full task lifecycle from creation to result retrieval.
Create tasks — Submit a task in plain English to human experts across specializations: data scraping, software development, design, copywriting, and general research/data work.
List tasks — Retrieve a paginated list of all your tasks with their statuses and details.
Check task status — Poll a specific task's current status (DRAFT → AWAITING_APPROVAL → PROCESSING → COMPLETED/CANCELLED/FAILED) and view pricing details once scoped.
Approve tasks — Review the quoted price and approve a task to trigger a charge and start human expert work.
Cancel tasks — Cancel a task at any stage (note: costs are not refunded after approval).
Fetch final results — Retrieve the final result text/report from a completed task.
Browse all result drafts — Access a paginated history of all intermediate and final results for a task, from latest to oldest.
Download file artifacts — Download files (images, documents) produced by experts and save them locally, using artifact references found in task results.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Tendem MCP delegate the task of researching competitors in the AI automation space"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Delegate tasks to vetted human experts — straight from your AI assistant. 🧑🔬🤝🤖
Tendem gives your AI direct access to vetted human experts for the work an agent alone can't nail: judgment-heavy research, competitive analysis, copywriting and editing, design review, data cleaning, and complex multi-step tasks. You (or your agent) submit a task in plain English — Tendem scopes the work, quotes a transparent price, and delivers verified results back into the chat as files, reports, or structured outputs. Agent plus expert raises the quality ceiling of what you can ship well beyond what either does alone.
Ideal for founders, creators, operators, and consultants who need reliable expert output without managing freelancers.
Get started → agent.tendem.ai.
This repo is the official distribution home for the Tendem MCP plugin — the part that lives in your AI client. Install it and your assistant arrives already knowing how to drive the full Tendem lifecycle: create → scope → approve → fetch, with clean polling and spend safety built in.
The MCP server itself is hosted at
https://mcp.tendem.ai/mcp— it is not in this repo. This repo contains the client-side plugins, skills, and rules that teach your AI client to use the hosted server well.
Sign in at agent.tendem.ai to create your account, top up your balance, and watch tasks run end-to-end. The full output of every task — markdown plus downloadable files — is always available there, no matter which client you launched it from.
The hosted server URL is the same everywhere: https://mcp.tendem.ai/mcp
(streamable HTTP, OAuth on first use). Pick your client:
/plugin marketplace add Toloka/tendem-mcp
/plugin install tendem@tendem-mcpcodex plugin marketplace add Toloka/tendem-mcp
codex plugin add tendem@tendem-mcpComplete the MCP server's OAuth login on first tool use.
copilot plugin marketplace add Toloka/tendem-mcp
copilot plugin install tendem@tendem-mcpgemini extensions install https://github.com/Toloka/tendem-mcpThen authenticate inside Gemini with /mcp auth tendem. The catalog entry is
gemini-extension.json at the repo root.
Kiro Powers register through the IDE. Point Kiro at the
tendem-power/ steering bundle in this repo, then connect the
tendem MCP server when prompted.
Add to your client's MCP config and sign in when prompted:
{
"mcpServers": {
"tendem": {
"type": "http",
"url": "https://mcp.tendem.ai/mcp"
}
}
}Related MCP server: human-delegation
For headless, CI, or programmatic use — where the interactive OAuth flow isn't viable — authenticate with a Tendem API key instead: sign in at agent.tendem.ai/mcp and create one under the Agent builders tab.
Python / LangChain / LangGraph: skip the raw MCP wiring and use
langchain-tendem — four price-capped
tools that let an agent create a task, answer the service's questions, and
collect the verified result, with all polling in plain Python (no LLM, no
tokens).
For every other stack, a native "type": "http" remote server triggers the
interactive OAuth flow, so
to pass a static API key instead, bridge the connection through
mcp-remote — a local stdio process
that connects to the hosted URL and injects your Authorization: ApiKey header:
{
"mcpServers": {
"tendem": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.tendem.ai/mcp",
"--transport",
"http-only",
"--header",
"Authorization: ApiKey <your-tendem-api-key>"
]
}
}
}Keep the token out of source control (inject it from an environment variable or
secret store). Your custom agent loop drives the same eleven tools the plugin
uses — create_task, get_task, get_contract, get_task_result,
list_tasks, approve_task, cancel_task, send_message, read_chat,
get_account, get_file_upload_url — to implement the create → scope →
approve → poll → fetch lifecycle yourself. The server ships a
tendem-quickstart prompt with a full walkthrough.
Spend safety:
approve_taskis what triggers a charge. In an automated pipeline, gate it behind your own approval logic and checkget_accountfirst — an insufficient balance returns a task-bound top-up URL, not an error to retry.
Path | What it is |
The plugin for Claude Code, Cursor, and GitHub Copilot CLI — skill, slash commands, background watcher agent, notification hook, and the MCP connector | |
The plugin for OpenAI Codex / ChatGPT — explicit | |
The Kiro Power (steering + MCP config) | |
The Gemini CLI extension manifest (root install) | |
This repo's Claude Code marketplace | |
This repo's Codex marketplace | |
MCP Registry manifest (remote server) | |
LLM discovery index |
Installing does more than wire up the connector — it ships steering so your AI handles Tendem correctly out of the box:
tendem-tasks skill — the end-to-end path (create → scope → approve →
fetch), silent long-polling without busy-loops, scope negotiation, and
live-verified file upload/download mechanics. Loads only when work actually
involves Tendem.
/tendem-task, /tendem-status, /tendem-result — explicit entry points
to submit, check, and fetch a task (in Codex these are $tendem-task etc.).
tendem-watcher agent (Claude / Copilot) — a background watcher that polls
a running task at a slow pace and pings you the moment it needs you.
Notification hook — a desktop notification when a task transitions to needing you (quote ready, top-up needed, input needed, result ready).
AGENTS.md — guide for AI agents working in this repo
CONTRIBUTING.md — how to contribute
Per-client manifests: Claude Code, OpenAI Codex, Cursor, GitHub Copilot CLI, and Gemini CLI
Product: tendem.ai · App: agent.tendem.ai
Legal: Privacy Policy · Terms of Use
Support: Help center
Tendem MCP is part of the Model Context Protocol ecosystem.
Approve a Tendem task and its price. A human expert will begin working (may take hours).
Call after reviewing the price in AWAITING_APPROVAL status. A human expert will then work on the task until it reaches COMPLETED status (may take hours).
Args: task_id: The Tendem task ID (UUID) to approve.
Returns: Confirmation message.
| Name | Required | Description | Default |
|---|---|---|---|
| task_id | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries full burden. It discloses that approval triggers human expert work and may take hours, which is a key behavioral trait. However, it does not mention any side effects, reversibility, or error states beyond the confirmation.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise, with a clear opening sentence followed by usage context in a separate paragraph, and standard Args/Returns sections. No unnecessary information is present.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple tool with one parameter and an output schema, the description adequately covers what the tool does, when to use it, and the parameter. It could mention idempotency or error conditions but is largely complete given the tool's complexity.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The description explains the task_id parameter as 'The Tendem task ID (UUID) to approve.', adding semantic detail beyond the schema's type definition. This helps the agent understand the expected format and purpose.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action 'approve' and the resource 'Tendem task and its price', distinguishing it from sibling tools like cancel_task, create_task, etc. It also specifies the resulting behavior (human expert begins working).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description tells when to use the tool: after reviewing the price in AWAITING_APPROVAL status. It also explains the expected outcome (human works for hours). However, it does not explicitly mention when not to use or provide alternative tools, but the context is clear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Cancel a Tendem task. Costs are not refunded after approval.
Can be called at any time. Note: costs are not refunded if cancelled after approval.
Args: task_id: The Tendem task ID (UUID) to cancel.
Returns: Confirmation message.
| Name | Required | Description | Default |
|---|---|---|---|
| task_id | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description must cover behavioral traits. It warns that costs are not refunded after approval, which is critical. However, it does not disclose other side effects like state changes, reversibility, or permissions needed.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise, with a clear first sentence stating the action and key warning. It then indicates usage timing and provides structured args/returns. No unnecessary words.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool has one parameter and no output schema, the description covers the main points: action, cost caveat, parameter, and return type. It lacks details about edge cases (e.g., already cancelled) but is sufficient for standard use.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, but the description adds meaningful information by specifying the parameter as 'Tendem task ID (UUID)' and stating its purpose ('to cancel'). This compensates for the lack of schema-level descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action 'Cancel' and the resource 'Tendem task'. It differentiates from sibling tools like approve_task and create_task by specifying the cancel operation.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description states 'Can be called at any time,' giving a usage context, but does not explicitly say when to avoid using it or compare to alternatives. The cost warning implies caution after approval, but no direct guidance about alternatives.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Create a new Tendem task for a human expert.
Poll get_task until AWAITING_APPROVAL to see the price.
After creation, poll with get_task until status is AWAITING_APPROVAL to see the price (may take up to 10 minutes).
When creating a task consider available human expert specializations: 1. data_scraping
Task types: High-volume data extraction, scraping, cleansing and enrichment (web, social media, or document sets).
Threshold: MUST be used when manual effort would exceed ~4 hours OR when tools like Selenium/BS4/Apify are required. 2. software_development
Task types: Debugging/refactoring existing code. Writing automation scripts. Building full-stack apps (Python, Node, TS). Building/adding features to WordPress, Woocommerce, Shopify based websites and stores. 3. design
Task types: Logos, Branding, Presentations (decks), Print Materials (flyers/brochures/billboards), and Packaging. 4. copywriting
Definition: Writing where style, tone, and usage of language is important.
Task types: SEO Writing, Newsletters, Press Releases, Case Studies, Ad Copy, Landing Page texts, Social Media posts, UX Writing, Email campaigns, Proofreading, editing, refining and humanizing ai text. 5. general
Definition: Expert level knowledge not required, good at attention to detail, using software and ai tools.
Task types: Manual data collection, enrichment, cleaning and analysis (incl. lead generation, contact list building), market research, formatting documents, converting files.
Avoid tasks that: Required Regulated Expertise: Medical diagnosis, legal advice, PhD-level research, or real-money investment advice. Required access to private/internal systems without providing credentials e.g., "Check my email"
Args: text: The task description/prompt to execute.
Returns: The created task.
| Name | Required | Description | Default |
|---|---|---|---|
| text | Yes |
| Name | Required | Description |
|---|---|---|
| task_id | Yes | |
| name | Yes | |
| status | Yes | |
| created_at | Yes | |
| approval_request_info | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries full burden. It explains polling behavior and creation time, but does not cover all side effects. Still, it adds significant behavioral context.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is well-structured with a clear first sentence, followed by polling instructions, specializations, and avoidance guidelines. It is appropriately sized for the tool's complexity, though slightly verbose.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the complexity of task creation with asynchronous polling and multiple specialization options, the description covers creation, polling, specializations, and prohibitions. An output schema exists, so return values are not needed.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The only parameter 'text' is described as 'The task description/prompt to execute,' adding meaning beyond the bare schema (type string, no description). With 0% schema coverage, description compensates well.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'Create a new Tendem task for a human expert,' specifying the verb and resource, and distinguishes from sibling tools like approve_task and cancel_task.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly lists when to use this tool with specializations and thresholds, and provides avoidance guidelines for regulated expertise or private system access, offering clear when-to-use and when-not-to-use advice.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Download a file artifact (image, document) from Tendem task results and save locally.
Artifact references appear in canvas content as:
aba://<artifact_id>Args: task_id: The Tendem task ID (UUID). artifact_id: The artifact ID (UUID) from the agents-reference block. path: The file path where the artifact should be saved.
Returns: Confirmation message with the saved file path and size.
| Name | Required | Description | Default |
|---|---|---|---|
| task_id | Yes | ||
| artifact_id | Yes | ||
| path | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Without annotations, the description carries the full burden. It discloses that the tool downloads and saves files locally and returns a confirmation with path and size. However, it omits details like whether existing files are overwritten or any permission requirements, leaving some behavioral gaps.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is very concise: a single sentence for purpose, a code block for artifact reference format, and clear sections for args and returns. Every sentence adds value with no redundancy.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The tool is simple with three required parameters and an output schema. The description covers purpose, usage, parameters, and return format. It lacks potential error handling or overwrite behavior, but overall is sufficient for the task's complexity.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema has no descriptions (0% coverage). The description compensates fully by explaining each parameter: task_id (UUID), artifact_id (UUID from agents-reference), and path (local save location). It adds context for artifact_id's origin, which is not evident from the schema alone.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool downloads a file artifact and saves it locally, specifying the source (Tendem task results) and file types (image, document). This distinctly differentiates it from sibling tools which focus on task management (approve, cancel, list, etc.).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explains how artifact references appear (agents-reference) and provides parameter details, giving clear context on when to use the tool. However, it does not explicitly compare with alternatives like get_task_result or state when not to use this tool, preventing a score of 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get all Tendem task results including intermediate drafts, from latest to oldest.
Args: task_id: The Tendem task ID (UUID) to get results for. page_number: Page number (0-indexed). page_size: Number of results per page (1-100).
Returns: Paginated Tendem task results with canvas content.
| Name | Required | Description | Default |
|---|---|---|---|
| task_id | Yes | ||
| page_number | Yes | ||
| page_size | Yes |
| Name | Required | Description |
|---|---|---|
| results | Yes | |
| total | Yes | |
| page_number | Yes | |
| page_size | Yes | |
| pages | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations provided, so description carries full burden. It discloses ordering (latest to oldest), inclusion of intermediate drafts, and pagination behavior. However, it does not mention whether the operation is read-only, authentication requirements, or any constraints beyond pagination.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise: a single-sentence summary followed by structured Args/Returns. It is front-loaded with the core purpose. Every sentence adds value, though the Returns section is brief.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The description explains pagination and ordering, and mentions 'canvas content' in returns. However, it does not clarify differences from sibling tools like 'get_task_result' or describe edge cases (e.g., no results). Output schema exists but is not shown, so evaluation is limited.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, but the description includes an Args block that explains each parameter: task_id is a UUID, page_number is 0-indexed, page_size is 1-100. This adds crucial meaning beyond the plain schema types, compensating for the lack of schema descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'Get all Tendem task results including intermediate drafts, from latest to oldest.' It specifies the verb (get), resource (task results), and scope (all, including drafts). Although it shares a server with 'get_task_result' (singular), the plural name and mention of 'all' and 'intermediate drafts' distinguishes it, but explicit differentiation is missing.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance on when to use this tool versus alternatives like 'get_task_result' or other siblings. It does not mention use cases or prerequisites.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get Tendem task status and details. Use to poll after create_task or approve_task.
Use to poll task status. After create_task, wait for AWAITING_APPROVAL to see price. After approve_task, a human expert works on the task until COMPLETED (may take hours).
Args: task_id: The Tendem task ID (UUID) to get.
Returns: The Tendem task including status and approval info if awaiting approval.
| Name | Required | Description | Default |
|---|---|---|---|
| task_id | Yes |
| Name | Required | Description |
|---|---|---|
| task_id | Yes | |
| name | Yes | |
| status | Yes | |
| created_at | Yes | |
| approval_request_info | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations provided, so description carries full burden. It describes the polling behavior and states but omits potential side effects or rate limits; adequate for a read tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Clear, well-structured with bullet points for args and returns. No redundant sentences, but could be slightly more concise.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given simple 1-parameter tool and presence of output schema, the description covers usage flow and return details adequately, though no info on error conditions.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema has 0% description coverage, but description adds that task_id is a UUID, providing crucial format context beyond the schema's type definition.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool retrieves task status and details, specifies the resource (Tendem task), and distinguishes it from sibling tools like approve_task, cancel_task, etc.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly states 'Use to poll after create_task or approve_task' and explains expected states (AWAITING_APPROVAL, COMPLETED), providing clear usage context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get the final result text from a completed Tendem task.
Args: task_id: The Tendem task ID (UUID).
Returns: The content of the latest canvas, or an error if task is not completed.
| Name | Required | Description | Default |
|---|---|---|---|
| task_id | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Given no annotations, the description reveals that the tool returns an error if the task is not completed and returns 'content of the latest canvas'. This is decent but could mention idempotency or read-only nature.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise with clear Args and Returns sections, no redundant information, and every sentence adds value.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The description covers the core functionality and error condition. With an output schema present, it does not need to detail return values further. Slight lack of usage context with siblings.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 0%, but the description adds the meaning of task_id as 'The Tendem task ID (UUID)', providing format and type beyond the schema's raw 'string' type.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'Get the final result text from a completed Tendem task', specifying the action, resource, and scope. It distinguishes itself from siblings like get_all_task_results and get_task.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies it should be used for completed tasks (returns error otherwise), but it does not explicitly compare with alternatives like get_task or get_all_task_results, nor provide when-not-to-use guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List all Tendem tasks with their statuses.
Args: page_number: Page number (0-indexed). page_size: Number of results per page (1-100).
Returns: Paginated list of Tendem tasks.
| Name | Required | Description | Default |
|---|---|---|---|
| page_number | Yes | ||
| page_size | Yes |
| Name | Required | Description |
|---|---|---|
| tasks | Yes | |
| total | Yes | |
| page_number | Yes | |
| page_size | Yes | |
| pages | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations provided, so the description carries the burden. It discloses pagination behavior via page_number and page_size, but lacks details on auth, rate limits, or data freshness. Adequate for a simple read operation.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Very concise, using a clear structured format (Args, Returns). Every sentence adds value with no wasted words.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With an output schema present, the description need not detail return values. It mentions 'Paginated list of Tendem tasks', which is sufficient. Could specify ordering or filtering but not essential for a list tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, but the description explains parameters: 'Page number (0-indexed)' and 'Number of results per page (1-100)'. This adds meaning beyond the schema types.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'List all Tendem tasks with their statuses', using a specific verb and resource. It distinguishes itself from sibling tools like get_task (single task) and create_task (creation).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance on when to use this tool versus alternatives such as get_task or get_all_task_results. The description does not mention any exclusions or context for selection.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.2approve_taskcancel_taskcreate_taskdownload_artifactget_all_task_resultsget_taskget_task_resultlist_tasksScored across 8 tools
Each tool targets a distinct action on tasks or artifacts: approve, cancel, create, download, get status, get final result, get all results, and list. No overlap in functionality.
All tool names follow a consistent verb_noun pattern (e.g., approve_task, download_artifact) using lowercase and underscores, making it easy to predict tool names.
With 8 tools, the set covers the full lifecycle of task management without being bloated or sparse. Each tool has a clear purpose.
The tools cover creation, approval, cancellation, status polling, result retrieval (single and all), artifact download, and listing. The only missing operation is updating a task, but that is not critical for the typical workflow.
Nifty's MCP server — exposes tasks, projects, messages, and files as tools for AI agents.
MCP Server for an Agent Task Marketplace
Official MCP server for Agentwork — delegate tasks to AI agents with human-in-the-loop
Hosted MCP server connecting AI assistants to 9,000+ apps and 40,000+ actions via Zapier.
