Server data from the Official MCP Registry
Huly MCP: feature-complete MCP server for Huly integration
About
Huly MCP: feature-complete MCP server for Huly integration
Security Report
Huly MCP is a Huly project-management integration that requires a Huly API token supplied via environment variables, which is consistent with its purpose. The visible code shows no exfiltration, backdoor, or shell injection, and the native stdio bridge and fallback diagnostic server appear carefully validated. Remaining concerns are limited to the truncated sections of the source, the optional HTTP transport exposure on 0.0.0.0, and the broad permissions inherent to a token-bearing integration with full account rights. Supply chain analysis found 2 known vulnerabilities in dependencies (0 critical, 1 high severity). Package verification found 1 issue.
10 files analyzed · 7 issues found
Security scores are indicators to help you make informed decisions, not guarantees. Always review permissions before connecting any MCP server.
Permissions Required
This plugin requests these system permissions. Most are normal for its category.
What You'll Need
Set these up before or after installing:
Environment variable: MCP_TRANSPORT
Environment variable: HULY_URL
Environment variable: HULY_EMAIL
Environment variable: HULY_PASSWORD
Environment variable: HULY_WORKSPACE
Environment variable: HULY_CONNECTION_TIMEOUT
Environment variable: HULY_TOOL_MODE
Environment variable: PROXY_OUTPUT_STRICT
Environment variable: TOOLSETS
Environment variable: TOOLS
How to Install
Add this to your MCP configuration file:
{
"mcpServers": {
"huly-mcp": {
"command": "npx",
"args": [
"-y",
"@firfi/huly-mcp@0.56.0"
],
"env": {
"TOOLS": "your-tools-here",
"HULY_URL": "your-huly-url-here",
"TOOLSETS": "your-toolsets-here",
"HULY_EMAIL": "your-huly-email-here",
"HULY_PASSWORD": "your-huly-password-here",
"MCP_TRANSPORT": "your-mcp-transport-here",
"HULY_TOOL_MODE": "your-huly-tool-mode-here",
"HULY_WORKSPACE": "your-huly-workspace-here",
"PROXY_OUTPUT_STRICT": "your-proxy-output-strict-here",
"HULY_CONNECTION_TIMEOUT": "your-huly-connection-timeout-here"
}
}
}
}Some servers take additional arguments. See the repository for details.
Documentation
View on GitHubFrom the project's GitHub README.
Huly MCP
Huly MCP is a feature-complete MCP server for Huly integration. Published on npm as @firfi/huly-mcp.
It connects to Huly Platform deployments and should also work with the Platform Collective and TraceX forks. Intabia 0.8.33 / model 0.8.4 has been compatibility-tested; known gaps are tracked in issue #279. Fork support remains unofficial.
Huly CLI Alternative
Looking for a Huly CLI, terminal-native Huly integration, or a scripting and automation alternative to running an MCP server? Huly CLI is a feature-complete command-line interface for Huly, published as @firfi/huly-cli, with full Huly operation parity with Huly MCP. It exposes the same Huly operations as native, schema-aware commands with human-readable or JSON output, file-backed inputs, uploads, and downloads.
Huly MCP and Huly CLI are maintained together in this monorepo and derive from the same shared operation registry, keeping operation inputs, outputs, warnings, and typed failures aligned. See the Huly CLI documentation for the complete generated command reference, or browse CLI examples, Huly automation patterns, and the Huly MCP vs CLI comparison.
Installation
Node.js requirement
Huly MCP requires Node.js 22.19.0 or later. MCP clients do not provide a Node.js runtime: node, npx,
and the #!/usr/bin/env node package executable resolve Node.js from the environment in which the MCP client
runs. Check that environment before installation:
node --version
If an MCP client starts Huly MCP with an older runtime, Huly MCP initializes a diagnostic-only MCP surface instead
of loading the incompatible application. Its server instructions and get_huly_startup_diagnostic tool report the
detected executable and required version, and the same unsupported Node.js runtime diagnostic is written to
stderr. Upgrade Node.js or configure the MCP server command to use the absolute path of a compatible Node.js
executable, then restart the MCP server. Huly operations remain unavailable until the runtime is upgraded.
The standard configuration uses a Huly API token and works with most MCP clients:
{
"mcpServers": {
"huly": {
"command": "npx",
"args": ["-y", "@firfi/huly-mcp@latest"],
"env": {
"MCP_TRANSPORT": "stdio-resident",
"HULY_URL": "https://your-huly-instance.example.com",
"HULY_TOKEN": "your-huly-api-token",
"HULY_WORKSPACE": "yourworkspace"
}
}
}
}
Prefer a managed API token created in Huly account settings when the target deployment supports it. Managed tokens are bound to one workspace, assigned an expiry, and can be revoked without changing the account password. They still inherit the account's full rights in that workspace, so use a dedicated integration account with only the workspace role, space memberships, and data access the automation needs. Older Huly deployments may not expose managed token settings; email/password authentication remains available as a compatibility fallback.
Tool Exposure Defaults
By default (HULY_TOOL_MODE=auto), Huly MCP optimizes for current MCP clients by avoiding a 470-tool eager list. Exact claude-code sessions receive native Huly tools. Codex, Cursor, Windsurf, Copilot, OpenCode, Claude AI/Desktop-style clients, and unknown clients receive a small proxy surface: list_tool_categories, search_tools, get_tool_schema, and invoke_tool.
Exact native tool names still dispatch when a client calls them directly, but many clients only call tools returned by tools/list. Set HULY_TOOL_MODE=native to make every Huly tool appear first-class, or use TOOLSETS / TOOLS to pin frequently used native tools while keeping proxy discovery available.
npx automatically downloads and caches Huly MCP; pnpm dlx @firfi/huly-mcp@latest is an alternative. Huly is optional, so Codex can start without it and a cold download may leave the first turn without Huly tools. Run /mcp to check the connection.
Add to ~/.codex/config.toml (the 120-second startup allowance includes installation):
[mcp_servers.huly]
startup_timeout_sec = 120
command = "npx"
args = ["-y", "@firfi/huly-mcp@latest"]
[mcp_servers.huly.env]
MCP_TRANSPORT = "stdio-resident"
HULY_URL = "https://your-huly-instance.example.com"
HULY_TOKEN = "your-huly-api-token"
HULY_WORKSPACE = "yourworkspace"
claude mcp add huly \
-e MCP_TRANSPORT=stdio-resident \
-e HULY_URL=https://your-huly-instance.example.com \
-e HULY_TOKEN=your-huly-api-token \
-e HULY_WORKSPACE=yourworkspace \
-- npx -y @firfi/huly-mcp@latest
Or add to ~/.claude.json using the standard config above.
Add the standard config to your claude_desktop_config.json:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
Add with Command Palette → "MCP: Add Server", or put this in a VS Code MCP config such as .vscode/mcp.json. Do not commit workspace config files that contain real credentials.
{
"servers": {
"huly": {
"command": "npx",
"args": ["-y", "@firfi/huly-mcp@latest"],
"env": {
"MCP_TRANSPORT": "stdio-resident",
"HULY_URL": "https://your-huly-instance.example.com",
"HULY_TOKEN": "your-huly-api-token",
"HULY_WORKSPACE": "yourworkspace"
}
}
}
}
Add the standard config to ~/.cursor/mcp.json, or via Settings → Tools & Integrations → New MCP Server.
Add the standard config to your Windsurf MCP configuration file.
Open the global configuration file (~/.config/opencode/opencode.json) and merge this entry into your config:
{
"mcp": {
"huly": {
"type": "local",
"command": ["npx", "-y", "@firfi/huly-mcp@latest"],
"environment": {
"MCP_TRANSPORT": "stdio-resident",
"HULY_URL": "https://your-huly-instance.example.com",
"HULY_TOKEN": "your-huly-api-token",
"HULY_WORKSPACE": "yourworkspace"
}
}
}
}
Pi requires an MCP adapter. Install it, then use Huly MCP in auto mode with the four proxy tools
promoted into Pi's native tool list:
pi install npm:pi-mcp-adapter
See the Pi setup guide for the user-global configuration, optional
TOOLSETS/TOOLS native pins, and verification steps.
Updating
The @latest tag asks the package runner for the newest version. Some MCP clients keep server processes or resolved installs alive, so restart or re-add the server when updating:
| Client | How to update |
|---|---|
| Codex | codex mcp remove huly then re-add with the install command above. If your password has shell-sensitive characters, edit ~/.codex/config.toml directly instead |
| Claude Code | claude mcp remove huly then re-add with the install command above |
| Claude Desktop | Restart the app (it runs npx on startup) |
| VS Code / Cursor | Restart the MCP server from the command palette/configured client or reload the window |
| OpenCode | Restart OpenCode or start a new session after config changes |
| Pi | Run /mcp reconnect huly, then /reload if the registered direct-tool surface changed |
| npx (manual) | MCP_TRANSPORT=stdio-resident npx -y @firfi/huly-mcp@latest — the -y flag auto-confirms install prompts |
Maintainer Production Release
Publish from a clean, current master checkout with the preferred certified Node.js runtime:
mise exec node@24.15.0 -- pnpm local-release
Use this exact command even when a newer Node.js version is active on the host. The release build accepts only the exact Node.js versions certified by CI; this is intentionally narrower than the package's runtime requirement. See the production release guide for preflight checks, rerun behavior, and verification.
Shared resident stdio
The setup examples explicitly select MCP_TRANSPORT=stdio-resident. MCP clients still use stdio, while compatible clients under the same OS user share a Node backend through small native bridges. Linux and macOS arm64/x64 binaries are included; no Rust installation is needed. Application requests remain buffered during the 30-second resident startup budget. Startup failure launches ordinary stdio and dispatches the buffered application input once. A failed active resident never reconnects or replays application requests.
Omitting MCP_TRANSPORT still selects ordinary stdio, preserving existing installations. Set MCP_TRANSPORT=stdio explicitly to keep a separate backend per client. Restart the MCP client after changing the setting.
Manual resident backend
The experimental manual resident backend path lets independent stdio hosts select one explicitly started local backend through a private attachment file. Full qualification remains pending.
HTTP Transport
By default, the server uses stdio transport. For HTTP transport:
HULY_URL=https://your-huly-instance.example.com \
HULY_TOKEN=your-huly-api-token \
HULY_WORKSPACE=yourworkspace \
MCP_TRANSPORT=http \
npx -y @firfi/huly-mcp@latest
Server listens on http://127.0.0.1:3000/mcp by default.
Configure with MCP_HTTP_PORT and MCP_HTTP_HOST:
MCP_TRANSPORT=http MCP_HTTP_PORT=8080 MCP_HTTP_HOST=0.0.0.0 npx -y @firfi/huly-mcp@latest
HTTP MCP Protocol Support
Both transports are implemented with Effect AI's MCP server and protocol adapters. The server configures the released MCP 2026-07-28 stateless protocol and the MCP 2025-06-18 compatibility protocol. Modern HTTP requires one JSON-RPC message per POST, Accept: application/json, text/event-stream, MCP-Protocol-Version: 2026-07-28, Mcp-Method, method-specific Mcp-Name, and per-request _meta.io.modelcontextprotocol/* client metadata. Legacy HTTP supports initialized sessions using the returned Mcp-Session-Id and bare stateless calls with request-local bindings. Huly credentials are configured separately through env vars or supported x-huly-* headers.
See MCP transport architecture and certification for the implementation boundary and the current verification record.
For hosted or tunneled HTTP deployments, you can require an MCP endpoint bearer token:
MCP_TRANSPORT=http \
MCP_AUTH_TOKEN="$(openssl rand -hex 32)" \
npx -y @firfi/huly-mcp@latest
HTTP clients must then send:
Authorization: Bearer <MCP_AUTH_TOKEN>
MCP_AUTH_TOKEN protects only the MCP HTTP /mcp endpoint. It is unrelated to HULY_TOKEN, does not authenticate to Huly, and does not replace HULY_EMAIL / HULY_PASSWORD / HULY_TOKEN. Huly credentials are still required through process env vars or, for hosted URL deployments, the supported x-huly-* headers. Stdio deployments do not use MCP_AUTH_TOKEN.
Shared CLI and stdio profiles
Save a profile with huly auth login --profile work, then set MCP_TRANSPORT=stdio-resident and HULY_PROFILE=work in the stdio MCP server's environment. Both processes must run under the same OS user and configuration directory. Stdio never follows the active CLI profile implicitly. Saved tokens are bound to their instance URL and workspace; changes require login again or complete environment credential injection. See the CLI profile instructions. HTTP configuration remains unchanged.
Hosted HTTP Header Configuration
For hosted URL deployments, keep the server process configured with MCP_TRANSPORT=http. A hosting layer can forward per-session Huly credentials as request headers, so one hosted server can serve different Huly workspaces without process-wide HULY_* env vars.
Supported v1 headers:
| Header | Required | Description |
|---|---|---|
x-huly-url | Yes | Huly instance URL |
x-huly-workspace | Yes | Workspace identifier |
x-huly-token | Yes | Huly API token |
x-huly-connection-timeout | No | Connection timeout in ms |
If any supported x-huly-* header is present, all required headers must be present and unsupported x-huly-* headers are rejected. Requests with only unsupported x-huly-* headers (for example a proxy trace id) use the process environment configuration. Missing values are not filled from environment variables. Email/password auth is not supported in hosted header configuration v1.
For a Smithery publish schema example, see docs/SMITHERY_URL_PUBLISH.md.
Environment Variables
| Variable | Required | Description |
|---|---|---|
HULY_URL | Yes | Reachable self-hosted or replacement hosted Huly instance URL |
HULY_TOKEN | Auth* | Preferred: managed Huly API token where supported |
HULY_EMAIL | Auth* | Compatibility fallback: account email |
HULY_PASSWORD | Auth* | Compatibility fallback: account password |
HULY_WORKSPACE | Yes | Workspace identifier |
HULY_CONNECTION_TIMEOUT | No | Connection timeout in ms. Omit to use the package default. |
MCP_TRANSPORT | No | Transport type: stdio (package default), stdio-resident (recommended in setup examples), or http |
MCP_PRELOAD_TOOLS | No | Load every tool handler before serving requests. Defaults to false; use true for smoke tests that check all handler imports. |
MCP_HTTP_PORT | No | HTTP server port (falls back to PORT, then 3000) |
MCP_HTTP_HOST | No | HTTP server host. Omit to bind to the package default loopback host. |
MCP_AUTH_TOKEN | No | Optional bearer token required by HTTP clients for /mcp. This protects the MCP endpoint only; it is not a Huly API token. |
HULY_TOOL_MODE | No | Tool exposure mode: auto (default), native, or proxy. auto keeps exact claude-code native and resolves Codex, Cursor, Windsurf, Copilot, opencode, Claude AI, and unknown clients to proxy mode. |
PROXY_OUTPUT_STRICT | No | Proxy candidate strictness: false (default) keeps proxy discovery broad; true makes active TOOLSETS / TOOLS a hard allow-list for proxy search, schema lookup, and invocation. |
TOOLSETS | No | Comma-separated tool categories to expose. If neither TOOLSETS nor TOOLS is set, all native Huly tools are exposed. Example: issues,projects,search |
TOOLS | No | Comma-separated exact tool names to expose in addition to selected toolsets. Example: list_documents,create_issue |
HULY_MCP_TELEMETRY | No | Aggregate usage analytics (PostHog), on by default. Set 0 to disable. DO_NOT_TRACK=1 also disables it. |
*Auth: Prefer HULY_TOKEN. Use HULY_EMAIL + HULY_PASSWORD only when the Huly deployment does not support a
managed API token or another compatibility constraint requires password authentication.
Startup and initialization
Tool discovery reads a generated catalog without loading operation handlers or waiting for Huly. Handlers load by tool group on demand; concurrent calls share the group's load. Run pnpm update-tool-catalog after changing tool definitions. Builds reject a stale catalog.
For stdio, the first tool-list request starts the process-owned Huly connection in the background. Calls that need the full Huly model wait for the same initialization. Cancelling a waiting call does not cancel initialization for other calls. A failed model initialization can be retried by a later call. Project listing uses a separate REST read and does not wait for the full model. Other operations retain the full model and their existing behavior.
The default transport remains stdio: each process launched by an MCP host owns its connection and model. HTTP lets clients share one running server. Connections created from request credential headers initialize the model only when needed; the CLI also uses this on-demand policy.
Built-in Diagnostic Tools
get_version returns the current server version and latest npm version.
get_huly_context returns sanitized runtime/configuration context for the current MCP session without connecting to Huly. It reports package version, transport, auth mode, sanitized Huly URL origin/host/protocol, workspace, timeout, config sources, native tool scope filtering, and resolved native/proxy tool exposure. Tokens, passwords, email values, credential headers, URL paths, URL query strings, and URL credentials are never returned.
MCP Resources
The server exposes read-only MCP Resources as JSON context for clients that support resources/read.
| Template | Name | Description | MIME Type |
|---|---|---|---|
huly://projects/{project} | huly-project | Read full details for a Huly tracker project by project identifier, for example huly://projects/HULY. | application/json |
huly://issues/{issue} | huly-issue | Read full details for a Huly issue by full issue identifier, for example huly://issues/HULY-123. | application/json |
huly://projects/{project}/issues/{issue} | huly-project-issue | Read full details for a Huly issue by project identifier and issue number, for example huly://projects/HULY/issues/123. | application/json |
resources/list returns concrete active project resources. Issue resources are template-based: use resources/templates/list to discover supported issue URI templates, then read a known issue URI.
Backlog
Feature backlog and SDK parity notes live in docs/BACKLOG.md.
Available Tools
When resolved tool exposure is proxy, clients see the built-in tools plus these proxy meta-tools. Native Huly tools are then discovered and invoked through the proxy candidate catalog. Exact native tool names also dispatch when a client calls them directly, subject to PROXY_OUTPUT_STRICT scope rules, but hidden native tools are not advertised through tools/list.
Proxy Meta-Tools
| Tool | Description |
|---|---|
list_tool_categories | Lists Huly tool categories available through this proxy. Use this first when you need a broad map of capabilities before searching for a specific Huly tool. |
search_tools | Searches the current proxy-visible Huly tool catalog by tool name, category, description, and parameter names. Returns exact tool names plus required and optional parameter names for single-call follow-up with get_tool_schema or invoke_tool. |
get_tool_schema | Returns the exact input and output schema for one proxy-visible Huly tool. Use this before invoke_tool when you are not certain about required argument names or result shape. |
invoke_tool | Invokes one proxy-visible Huly tool by exact name with its arguments. This tool can call read or write Huly operations; check get_tool_schema and the target tool annotations when safety matters. |
TOOLSETS categories: projects, issues, comments, milestones, documents, storage, attachments, contacts, channels, calendar, time tracking, search, associations, activity, notifications, workspace, approvals, boards, cards, collaborators, custom-fields, drive, hr, inventory, labels, leads, mail, templates, model-administration, planner, preferences, processes, recruiting, sdk-discovery, security-administration, sequence-administration, spaces, support, tag-categories, tags, task-management, test-management, user-statuses, views, virtual-office, workbench, workflow-statuses
Projects
| Tool | Description |
|---|---|
list_project_target_preferences | List low-level per-project tracker target preference records. These Huly ProjectTargetPreference records are attached to projects and used by tracker UI/workflows to remember target-related preference props. Omit project to list recent preferences across projects, or pass a project identifier to inspect one project's preference. Props are SDK-open key/value payloads. |
upsert_project_target_preference | Create or update the low-level ProjectTargetPreference record for a project. This refreshes usedOn and merges SDK-open target preference props by key. Use for tracker SDK parity or advanced administration; ordinary project and issue workflows usually do not need this tool. |
list_projects | List all Huly projects. Returns projects sorted by name. Supports filtering by archived status. |
get_project | Get full details of a Huly project including its statuses. Returns project name, description, archived flag, default status, and all available statuses. |
list_statuses | List all issue statuses for a Huly project with workflow category and default info. Returns status name, category, and isDefault. Use this to discover valid statuses before creating or updating issues. |
create_project | Create a new Huly tracker project. Idempotent: returns existing project if one with the same identifier already exists (created=false). Identifier must be 1-5 uppercase alphanumeric chars starting with a letter. |
update_project | Update a Huly project. Only provided fields are modified. Set description to null to clear it. |
delete_project | Permanently delete a Huly project. All issues, milestones, and components in this project will be orphaned. This action cannot be undone. |
Issues
| Tool | Description |
|---|---|
preview_deletion | Preview the impact of deleting a Huly entity before actually deleting it. Shows affected sub-entities, relations, and warnings. Supports issues, projects, components, and milestones. Use this to understand cascade effects before calling a delete operation. |
list_external_tracker_targets | List external issue-tracker targets mapped to a Huly project. Omit provider to list every provider supported by this server, or pass provider:'github' to list mapped GitHub repositories. Returns each target's provider, kind, stable Huly ID, exact name, enabled state, and an actionable unavailable reason when known. This discovers Huly configuration only; it does not call the external provider or prove that the synchronization worker is running. |
publish_issue_to_external_tracker | Request publication of an existing Huly issue as a new issue in Huly's configured external tracker. Identify the issue with project and its human identifier (for example, HULY-123); provider currently accepts only 'github'. Select target by stable Huly target ID or exact name, or omit target only when exactly one enabled target is mapped to the project. Publication is asynchronous: pending means Huly accepted the request, not that an external issue exists. Repeating the same target is safe while pending or published; a failed same-target request retries, while a different target is rejected. Use get_issue_publication_status to observe completion. |
get_issue_publication_status | Read the Huly-persisted external publication state for an issue identified by project and human issue identifier. Returns not_requested, pending, published, or failed; pending includes the ISO state-change timestamp and elapsed time, while published includes an external URL and positive issue number. A timeout is projected as failed only after the bounded wait, and later persisted publication takes precedence. This reads Huly state and never calls the external provider directly. |
list_components | List components in a Huly project. Components organize issues by area/feature. Returns components sorted by modification date (newest first). |
get_component | Retrieve full details for a Huly component. Use this to view component content and metadata. |
create_component | Create a new component in a Huly project. Components help organize issues by area/feature. Description supports markdown. Markdown links to current-workspace Huly browse URLs with _class, _id, and label become native Huly references; external URLs and other-workspace browse URLs stay normal links. Returns the created component ID and label. |
update_component | Update fields on an existing Huly component. Only provided fields are modified. Description supports markdown. Markdown links to current-workspace Huly browse URLs with _class, _id, and label become native Huly references; external URLs and other-workspace browse URLs stay normal links. |
set_issue_component | Set or clear the component on a Huly issue. Pass null for component to clear it. |
delete_component | Permanently delete a Huly component. This action cannot be undone. |
list_issues | Query Huly issues with optional filters. Returns issues sorted by modification date (newest first). Supports filtering by project, exact workflow status name (status), Huly SDK task.statusCategory key (statusCategory: UnStarted, ToDo, Active, Won, Lost), assignee (email, Person display name, or exact agent UserProfile title), creator (raw Person ID, exact email, or exact display name), component, a human-readable attached label title (label, exact and case-insensitive), milestone (project-scoped ID or exact case-insensitive label), milestone presence (hasMilestone), parentIssue (to list children of a specific issue), and isTopLevel (to return only native top-level issues). Creator, label, and milestone filters are applied before the result limit; ambiguous creator display names fail actionably and unknown creators return an empty list. Each result includes creator as a stable Person reference { id, name?, email? } when resolvable and the assigned milestone as { id, label } when resolvable. Supports searching by title substring (titleSearch) and description content (descriptionSearch). Each result includes deterministic label summaries with title and available color; missing labels are an empty array, duplicate titles collapse case-insensitively while preferring a valid color, and unusable partial attachments are omitted with a tool warning. |
get_issue | Retrieve full details for a Huly issue including markdown description, its creator as a stable Person reference { id, name?, email? } when resolvable, its assigned milestone as { id, label } when resolvable, and deterministic attached-label summaries with title and available color. Missing labels are an empty array, duplicate titles collapse case-insensitively while preferring a valid color, and unusable partial creator or label metadata is omitted with one bounded tool warning. Use this to view issue content, comments, classification, or full metadata. identifier accepts a project-local identifier/number or stable issue ID; stable IDs resolve across the workspace after movement and return the actual project. The supplied project must exist. |
create_issue | Create a new issue in a Huly project. Assignee accepts an email, Person display name, or exact agent UserProfile title and always stores the linked Person. Omit parentIssue for a native top-level issue, or specify parentIssue to create a sub-issue. Optionally set taskType by ID or display name; it is resolved within the target project's project type, and status is validated against that task type's workflow. Use list_task_types or get_project_type to discover valid task types and statuses. Description supports markdown formatting; markdown links to current-workspace Huly browse URLs with _class, _id, and label become native references, while external URLs stay normal links. Returns the created issue identifier. |
update_issue | Update fields on an existing Huly issue. Assignee accepts an email, Person display name, or exact agent UserProfile title and always stores the linked Person; use null to unassign. Optionally set taskType by ID or display name; it is resolved within the target project's project type, and the status is preserved only when valid for the new task type. Use list_task_types or get_project_type to discover valid task types and statuses. Only provided fields are modified. Description updates support markdown; markdown links to current-workspace Huly browse URLs with _class, _id, and label become native references, while external URLs stay normal links. |
add_issue_label | Add a tag/label to a Huly issue. Creates the tag if it doesn't exist in the project. |
remove_issue_label | Remove a tag/label from a Huly issue. Detaches the label reference; does not delete the label definition. |
delete_issue | Permanently delete a Huly issue. This action cannot be undone. |
move_issue | Move an issue tree within its project, preserving IDs, identifiers and attributes. issue/parent accept full identifiers or stable IDs; project accepts identifier or stable ID. destination forms: {project: 'HULY'} for project top level, {parent: 'HULY-42'} to infer the parent project, {project: 'HULY', parent: 'HULY-42'} requiring agreement, or {parent: null} for current project top level. Cross-project movement supports complete compatible trees with equal project types and actual supported kind/status, component/milestone references preserved when valid or matched by literal unique name; otherwise blocked with candidates for exact per-task resolutions {issueId,field,from,to}, where to is a valid destination ID or null to clear. All discoverable conflicts are returned before writes. Comments (including threads/nested attachments), issue attachments, labels, time reports and audited activity/replies/reactions follow the move; historical payloads and independent links stay intact. Unknown record classes/collection edges or discovery limits cause a pre-write refusal. Restricted projects are refused; private destinations require membership. Tree discovery follows attachments, and all task conflicts and resolutions are checked before number allocation. Consent never extends to siblings or newly discovered children. Safety limits refuse a complete move before writes; successful mappings include every task. Cross-project identifiers change; stable IDs and independent references remain. Omit resolutions for same-project calls, including no-ops. Inspects hierarchy before writes and verifies after execution; inconsistent or uncertain outcomes include inspection guidance. Never automatically retry an uncertain move. |
list_issue_templates | List issue templates in a Huly project. Templates define reusable issue configurations. Returns templates sorted by modification date (newest first). |
get_issue_template | Retrieve full details for a Huly issue template including children (sub-task templates). Use this to view template content, default values, and child template IDs. |
create_issue_template | Create a new issue template in a Huly project. Templates define default values for new issues. Template and child descriptions support markdown. Markdown links to current-workspace Huly browse URLs with _class, _id, and label become native Huly references; external URLs and other-workspace browse URLs stay normal links. Optionally include children (sub-task templates) that will become sub-issues when creating issues from this template. Returns the created template ID and title. |
create_issue_from_template | Create a new issue from a template. Applies template defaults, allowing overrides for specific fields. Description override supports markdown. Markdown links to current-workspace Huly browse URLs with _class, _id, and label become native Huly references; external URLs and other-workspace browse URLs stay normal links. If the template has children (sub-task templates), sub-issues are created automatically unless includeChildren is set to false. Returns the created issue identifier and count of children created. |
update_issue_template | Update fields on an existing Huly issue template. Only provided fields are modified. Description supports markdown. Markdown links to current-workspace Huly browse URLs with _class, _id, and label become native Huly references; external URLs and other-workspace browse URLs stay normal links. |
delete_issue_template | Permanently delete a Huly issue template. This action cannot be undone. |
add_template_child | Add a child (sub-task) template to an issue template. The child defines default values for sub-issues created when using create_issue_from_template. Child description supports markdown. Markdown links to current-workspace Huly browse URLs with _class, _id, and label become native Huly references; external URLs and other-workspace browse URLs stay normal links. Returns the child template ID. |
remove_template_child | Remove a child (sub-task) template from an issue template by its child ID. Get child IDs from get_issue_template response. |
add_issue_relation | Add a relation between two issues. Relation types: 'blocks' (source blocks target — pushes into target's blockedBy), 'is-blocked-by' (source is blocked by target — pushes into source's blockedBy), 'relates-to' (bidirectional link — updates both sides). targetIssue accepts cross-project identifiers like 'OTHER-42'. No-op if the relation already exists. |
remove_issue_relation | Remove a relation between two issues. Mirrors add_issue_relation: 'blocks' pulls from target's blockedBy, 'is-blocked-by' pulls from source's blockedBy, 'relates-to' pulls from both sides. No-op if the relation doesn't exist. |
list_issue_relations | List all relations of an issue. Returns blockedBy (issues blocking this one), blocks (issues this one blocks), relations (bidirectional issue links), and documents (linked documents with title/teamspace). |
link_document_to_issue | Link a Huly document to an issue. The link appears in the issue's Relations panel in the UI. Idempotent: no-op if the document is already linked. Use list_issue_relations to see linked documents. |
unlink_document_from_issue | Remove a document link from an issue. Idempotent: no-op if the document is not linked. |
list_related_issue_targets | List rules that choose the default destination project for related issues. A spaceRule says related issues from one space default to targetProject. A classRule says related issues for one object class default to targetProject. targetProject is a project identifier, or null for no default destination project. |
set_related_issue_target | Set the default destination project for related issues from a space or object class. For space, creates or updates a spaceRule. For objectClass, only updates an existing classRule; this tool never creates classRule targets. Pass targetProject as a project identifier, or null to clear the default destination project. |
delete_related_issue_space_target | Delete the spaceRule that chooses the default destination project for related issues from one space. This only deletes spaceRule targets; classRule deletion is intentionally unsupported because class rules may be model-provided. |
Comments
| Tool | Description |
|---|---|
list_comments | List comments on a Huly issue. Returns comments sorted by creation date (oldest first). |
add_comment | Add a comment to a Huly issue. Comment body supports markdown formatting. Markdown links to current-workspace Huly browse URLs with _class, _id, and label become native Huly references; external URLs and other-workspace browse URLs stay normal links. |
update_comment | Update an existing comment on a Huly issue. Comment body supports markdown formatting. Markdown links to current-workspace Huly browse URLs with _class, _id, and label become native Huly references; external URLs and other-workspace browse URLs stay normal links. |
delete_comment | Delete a comment from a Huly issue. This action cannot be undone. |
Milestones
| Tool | Description |
|---|---|
list_milestones | List milestones in a Huly project. Returns milestones sorted by modification date (newest first). |
get_milestone | Retrieve full details for a Huly milestone. Use this to view milestone content and metadata. |
create_milestone | Create a new milestone in a Huly project. Description supports markdown. Markdown links to current-workspace Huly browse URLs with _class, _id, and label become native Huly references; external URLs and other-workspace browse URLs stay normal links. Returns the created milestone ID and label. |
update_milestone | Update fields on an existing Huly milestone. Only provided fields are modified. Description supports markdown. Markdown links to current-workspace Huly browse URLs with _class, _id, and label become native Huly references; external URLs and other-workspace browse URLs stay normal links. |
set_issue_milestone | Set or clear the milestone on a Huly issue. Pass null for milestone to clear it. |
delete_milestone | Permanently delete a Huly milestone. This action cannot be undone. |
Documents
Documentation truncated — see the full README on GitHub.
Reviews
No reviews yet
Be the first to review this server!
More Developer Tools MCP Servers
Rea
Freeby Morluto · Developer Tools
Reverse engineer anything from your terminal or agent with one CLI and MCP server.
Paperclip
Freeby Paperclipai · Developer Tools
Trending hip-hop artist momentum scores across four cultural dimensions.
Scrapling
Freeby D4Vinci · Developer Tools
Web scraping with stealth HTTP, real browsers, and Cloudflare bypass. CSS selectors supported.
Toleno
Freeby Toleno · Developer Tools
Toleno Network MCP Server — Manage your Toleno mining account with Claude AI using natural language.
MCP Marketplace
Freeby mcp-marketplace · Developer Tools
Search and install MCP servers from inside your AI client.
mcp-creator-python
Freeby mcp-marketplace · Developer Tools
Create, build, and publish Python MCP servers to PyPI — conversationally.
