> ## Documentation Index
> Fetch the complete documentation index at: https://docs.useorgx.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# OpenClaw Plugin MCP Tools

> Code-generated reference for the OpenClaw plugin MCP bridge, including scoped domain availability.

This page documents the MCP tools exposed by the **OpenClaw plugin local bridge**, grounded in plugin source code.

> [!NOTE]
>
> Generated from OpenClaw plugin version `0.7.35`. Regenerate with `pnpm
>     docs:generate-openclaw-plugin`.

## Endpoint Model

| Endpoint | Purpose |
| - | - |
| `/orgx/mcp` | Full plugin tool surface (all 33 tools). |
| `/orgx/mcp/{domain}` | Domain-safe subset of tools. |

### Scoped Domains

| Domain | Tool Count | Additional Tools Beyond Base |
| - | - | - |
| `engineering` | 18 | None |
| `product` | 18 | None |
| `design` | 18 | None |
| `marketing` | 18 | None |
| `sales` | 18 | None |
| `operations` | 23 | `orgx_agent_sessions`, `orgx_apply_changeset`, `orgx_reassign_stream`, `orgx_reassign_streams`, `update_agent_config` |
| `orchestration` | 25 | `orgx_agent_sessions`, `orgx_apply_changeset`, `orgx_clear_agent_session`, `orgx_reassign_stream`, `orgx_reassign_streams`, `orgx_resume_agent_session`, `update_agent_config` |

## Tool Catalog

| Category | Tools |
| - | - |
| Sync & Reporting | `orgx_emit_activity`, `orgx_register_artifact`, `orgx_report_progress`, `orgx_status`, `orgx_sync` |
| Delegation & Assignment | `orgx_delegation_preflight`, `orgx_spawn_check` |
| Decision & Run Control | `orgx_checkpoint_restore`, `orgx_checkpoints_list`, `orgx_request_decision`, `orgx_run_action` |
| Planning & Entity Management | `orgx_apply_changeset`, `orgx_create_entity`, `orgx_reassign_stream`, `orgx_reassign_streams`, `orgx_update_entity`, `update_stream_progress` |
| Quality, Proof & Outcomes | `orgx_get_outcome_attribution`, `orgx_proof_status`, `orgx_quality_score`, `orgx_record_outcome`, `orgx_verify_completion` |
| Agent Policy & Configuration | `get_agent_config`, `list_agent_configs`, `orgx_sentinel_catalog`, `update_agent_config` |
| General | `orgx_agent_sessions`, `orgx_clear_agent_session`, `orgx_get_morning_brief`, `orgx_list_entities`, `orgx_query_org_memory`, `orgx_recommend_next_action`, `orgx_resume_agent_session` |

***

## Sync & Reporting

### orgx\_emit\_activity

Emit append-only OrgX activity telemetry (launch reporting contract primary write tool).

**initiative\_id** `string`

Initiative UUID (required unless ORGX\_INITIATIVE\_ID is set)

**message** `string` *required*

Human-readable activity update

**run\_id** `string`

Optional run UUID

**correlation\_id** `string`

Required when run\_id is omitted

**source\_client** `string`

Required when run\_id is omitted Options: `openclaw`, `codex`, `claude-code`,
`api`.

**phase** `string`

Reporting phase Options: `intent`, `execution`, `blocked`, `review`,
`handoff`, `completed`.

**progress\_pct** `number`

Optional progress percentage

**level** `string`

Optional level (default info) Options: `info`, `warn`, `error`.

**next\_step** `string`

Optional next step

**metadata** `object`

Optional structured metadata

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_register\_artifact

Register a work output (PR, document, config change, report, etc.) as a work\_artifact in OrgX. Makes it visible in the dashboard activity timeline and entity detail modals.

**initiative\_id** `string`

Convenience: initiative UUID. Used as entity\_type='initiative',
entity\_id=\<this> when entity\_type/entity\_id are not provided.

**entity\_type** `string`

The type of entity this artifact is attached to Options: `initiative`,
`milestone`, `task`, `decision`, `project`.

**entity\_id** `string`

UUID of the entity this artifact is attached to

**name** `string` *required*

Human-readable artifact name (e.g., 'PR #107: Fix build size')

**artifact\_type** `string` *required*

Artifact type code (e.g., 'eng.diff\_pack', 'pr', 'document'). Falls back to
'shared.project\_handbook' if the type is not recognized by OrgX.

**confidence\_score** `number`

Self-assessed confidence for this artifact in \[0,1].

**description** `string`

What this artifact is and why it matters

**url** `string`

Durable proof source for the artifact: a public permalink (GitHub
PR/commit/blob, published doc URL) or an absolute file path.

**content** `string`

Inline preview content (markdown/text). Supplemental only; it does not replace
a durable source URL or file path.

**metadata** `object`

Optional artifact metadata. Use for proof fields such as commit\_sha, branch,
artifact\_hash, quality\_gate, or schema validation signals.

**queue\_ref** `object`

Optional proof-chain queue reference linking this artifact to the
initiative/workstream/task queue item.

**run\_ref** `object`

Optional proof-chain run reference linking this artifact to the producing
run/session.

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_report\_progress

Alias for orgx\_emit\_activity. Report progress at key milestones so the team can track your work.

**initiative\_id** `string`

Initiative UUID (required unless ORGX\_INITIATIVE\_ID is set)

**run\_id** `string`

Optional run UUID

**correlation\_id** `string`

Required when run\_id is omitted

**source\_client** `string`

Options: `openclaw`, `codex`, `claude-code`, `api`.

**summary** `string` *required*

What was accomplished (1-2 sentences, human-readable)

**phase** `string` *required*

Current work phase Options: `researching`, `implementing`, `testing`,
`reviewing`, `blocked`.

**progress\_pct** `number`

Progress percentage (0-100)

**next\_step** `string`

What you plan to do next

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_status

Get current OrgX org status. Agents should pass their canonical agent\_id or domain to receive only runnable assigned tasks.

**agent\_id** `string`

Optional canonical agent ID, for example operations-agent

**domain** `string`

Optional domain filter Options: `engineering`, `product`, `design`,
`marketing`, `sales`, `operations`, `orchestration`.

**canonical\_only** `boolean`

When true, return only canonical next tasks

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_sync

Push/pull memory sync with OrgX. Send local memory/daily log; receive initiatives, tasks, decisions, model routing policy.

**memory** `string`

Local memory snapshot to push

**dailyLog** `string`

Today's session log to push

**agents** `object[]`

Optional local agent states to sync into OrgX

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

## Delegation & Assignment

### orgx\_delegation\_preflight

Run delegation preflight to score scope quality, estimate ETA/cost, and suggest a split before autonomous execution.

**intent** `string` *required*

Task intent in natural language

**acceptanceCriteria** `string[]`

Optional acceptance criteria to reduce ambiguity

**constraints** `string[]`

Optional constraints (deadline, stack, policy)

**domains** `string[]`

Optional preferred owner domains

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: Not exposed on scoped domain endpoints

***

### orgx\_spawn\_check

Check quality gate + get model routing before spawning a sub-agent. Returns allowed/denied, model tier, and check details.

**domain** `string` *required*

Agent domain (engineering, product, marketing, data, operations, design)

**taskId** `string`

OrgX task ID to check

**modelTier** `string`

Optional explicit model tier. Use standard for cheap functional validation;
omit to let OrgX auto-route. Options: `standard`, `balanced`, `precision`,
`local`, `sonnet`, `opus`.

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

## Decision & Run Control

### orgx\_checkpoint\_restore

Restore a run to a specific checkpoint.

**runId** `string` *required*

Run UUID

**checkpointId** `string` *required*

Checkpoint UUID

**reason** `string`

Optional restoration reason

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: Not exposed on scoped domain endpoints

***

### orgx\_checkpoints\_list

List checkpoints for a run.

**runId** `string` *required*

Run UUID

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: Not exposed on scoped domain endpoints

***

### orgx\_request\_decision

Alias for orgx\_apply\_changeset with decision.create. Request a human decision before proceeding.

**initiative\_id** `string`

Initiative UUID (required unless ORGX\_INITIATIVE\_ID is set)

**run\_id** `string`

Optional run UUID

**correlation\_id** `string`

Required when run\_id is omitted

**source\_client** `string`

Options: `openclaw`, `codex`, `claude-code`, `api`.

**question** `string` *required*

The decision question (e.g., 'Deploy to production?')

**context** `string`

Background context to help the human decide

**options** `string[]`

Available choices (e.g., \['Yes, deploy now', 'Wait for more testing',
'Cancel'])

**urgency** `string` *required*

How urgent this decision is Options: `low`, `medium`, `high`, `urgent`.

**blocking** `boolean`

Whether work should pause until this is decided (default: true)

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_run\_action

Apply a control action to a run: pause, resume, cancel, or rollback (rollback requires checkpointId).

**runId** `string` *required*

Run UUID

**action** `string` *required*

Control action Options: `pause`, `resume`, `cancel`, `rollback`.

**checkpointId** `string`

Checkpoint UUID (required for rollback)

**reason** `string`

Optional reason for audit trail

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: Not exposed on scoped domain endpoints

***

## Planning & Entity Management

### orgx\_apply\_changeset

Apply an idempotent transactional OrgX changeset (launch reporting contract primary mutation tool).

**initiative\_id** `string`

Initiative UUID (required unless ORGX\_INITIATIVE\_ID is set)

**idempotency\_key** `string`

Idempotency key (\<=120 chars). Auto-generated if omitted.

**operations** `object[]` *required*

Changeset operations (task.create, task.update, milestone.update,
decision.create)

**run\_id** `string`

Optional run UUID

**correlation\_id** `string`

Required when run\_id is omitted

**source\_client** `string`

Required when run\_id is omitted Options: `openclaw`, `codex`, `claude-code`,
`api`.

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: `operations`, `orchestration`

***

### orgx\_create\_entity

Create an OrgX entity (initiative, workstream, task, decision, milestone, etc.).

**type** `string` *required*

Entity type: initiative, workstream, task, decision, milestone, artifact,
agent, blocker

**title** `string` *required*

Entity title

**summary** `string`

Description

**status** `string`

Initial status (active, not\_started, todo)

**initiative\_id** `string`

Parent initiative ID (for workstreams/tasks)

**workstream\_id** `string`

Parent workstream ID (for tasks)

**workspace\_id** `string`

Workspace ID (canonical; preferred for new callers)

**command\_center\_id** `string`

Deprecated alias for workspace\_id

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: Not exposed on scoped domain endpoints

***

### orgx\_reassign\_stream

Reassign a workstream's stream ownership/agents and return reassignment scheduling details.

**workstream\_id** `string` *required*

Workstream UUID to reassign

**initiative\_id** `string` *required*

Parent initiative UUID

**status** `string`

Optional workstream status override (active, in\_progress, pending, etc.)

**domain** `string`

Optional target domain for the reassigned stream

**role** `string`

Optional role hint for assignment routing

**assigned\_agent\_ids** `string[]`

Optional assigned agent IDs

**assignedAgentIds** `string[]`

Alias for assigned\_agent\_ids

**assigned\_agent\_names** `string[]`

Optional assigned agent display names

**assignedAgentNames** `string[]`

Alias for assigned\_agent\_names

**assigned\_agents** `object[]`

Optional structured assigned agent list

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: `operations`, `orchestration`

***

### orgx\_reassign\_streams

Convenience batch reassignment tool. Takes initiative\_id and a workstream-to-domain mapping, then updates all listed workstreams.

**initiative\_id** `string` *required*

Parent initiative UUID for all workstream updates

**workstream\_domains** `object`

Mapping of workstream UUID -> target domain

**mapping** `object`

Alias for workstream\_domains: mapping of workstream UUID -> target domain

**mappings** `object[]`

Optional array mapping for advanced routing fields (workstream\_id, domain,
role, status).

**status** `string`

Optional workstream status override applied to each reassigned stream

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: `operations`, `orchestration`

***

### orgx\_update\_entity

Update an existing OrgX entity by type and ID.

**type** `string` *required*

Entity type

**id** `string` *required*

Entity UUID

**status** `string`

New status

**title** `string`

New title

**summary** `string`

New summary

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: Not exposed on scoped domain endpoints

***

### update\_stream\_progress

Legacy alias for orgx\_report\_progress. Report progress at key milestones so the team can track your work.

**initiative\_id** `string`

Initiative UUID (required unless ORGX\_INITIATIVE\_ID is set)

**run\_id** `string`

Optional run UUID

**correlation\_id** `string`

Required when run\_id is omitted

**source\_client** `string`

Options: `openclaw`, `codex`, `claude-code`, `api`.

**summary** `string` *required*

What was accomplished (1-2 sentences, human-readable)

**phase** `string` *required*

Current work phase Options: `researching`, `implementing`, `testing`,
`reviewing`, `blocked`.

**progress\_pct** `number`

Progress percentage (0-100)

**next\_step** `string`

What you plan to do next

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

## Quality, Proof & Outcomes

### orgx\_get\_outcome\_attribution

Get outcome attribution data for a task or run. Shows what value was attributed to the work and with what confidence. Required for L6 (Economics) proof level.

**task\_id** `string`

Task ID to get attribution for

**run\_id** `string`

Run ID to get attribution for

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_proof\_status

Check the proof chain status for a task or run. Returns a checklist of proof levels (L1-L7) with gaps highlighted. Use this to verify work is provably complete before marking done.

**task\_id** `string`

ID of the task to check proof for

**run\_id** `string`

ID of the run to check proof for

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_quality\_score

Record a quality score (1-5) for completed agent work. Used to gate future spawns and track performance.

**taskId** `string` *required*

ID of the completed task

**domain** `string`

Legacy alias for agentDomain

**agentDomain** `string`

Agent domain that did the work

**score** `number` *required*

Quality 1 (poor) to 5 (excellent)

**notes** `string`

Notes on the assessment

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_record\_outcome

Record an outcome event for a completed run, linking execution to business value. Required for L5 (Impact) proof level.

**initiative\_id** `string` *required*

Initiative ID this outcome belongs to

**execution\_id** `string` *required*

Execution/run ID that produced this outcome

**execution\_type** `string`

Execution type label. Defaults to agent\_run when omitted.

**run\_id** `string`

Existing OrgX run ID to attach the outcome to. If omitted, correlation\_id and
source\_client are used.

**correlation\_id** `string`

Idempotency/correlation ID for this execution. Defaults to execution\_id when
run\_id is omitted.

**source\_client** `string`

Client that produced the outcome. Defaults to openclaw when run\_id is omitted.
Options: `openclaw`, `codex`, `claude-code`, `api`.

**agent\_id** `string` *required*

Agent that did the work

**success** `boolean` *required*

Whether the execution was successful

**quality\_score** `number`

Quality score 1-5

**domain** `string`

Agent domain

**metadata** `object`

Additional outcome metadata

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_verify\_completion

Verify that an entity (task/milestone/workstream) meets completion requirements including proof chain. Returns structured result with any blocking issues.

**entity\_type** `string` *required*

Entity type: task, milestone, workstream

**entity\_id** `string` *required*

ID of the entity to verify

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

## Agent Policy & Configuration

### get\_agent\_config

Get a single agent behavior config by id (default only).

**config\_id** `string`

Agent config identifier. Use 'default'.

**refresh\_remote** `boolean`

When true, re-fetch behavior config policy from OrgX before returning data.

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### list\_agent\_configs

List available agent behavior configs managed by this plugin.

**refresh\_remote** `boolean`

When true, re-fetch behavior config policy from OrgX before returning data.

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_sentinel\_catalog

List built-in proactive sentinel templates. Supports optional domain filter.

**domain** `string`

Optional domain filter (engineering, sales, operations, product).

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: Not exposed on scoped domain endpoints

***

### update\_agent\_config

Update plugin-managed agent behavior config policy (default config only).

**config\_id** `string`

Agent config identifier. Use 'default'.

**frozen** `boolean`

Freeze remote skill-pack refresh when true.

**pinned\_checksum** `string`

Pin behavior policy to this exact checksum.

**pin\_to\_current** `boolean`

Pin to current local/remote checksum.

**clear\_pin** `boolean`

Clear any pinned checksum.

**action** `string`

Use 'rollback' to revert to the previous config version.

**rollback\_to\_audit\_id** `string`

Optional audit entry id to rollback to. When omitted with action='rollback',
uses the most recent audit entry.

**template\_id** `string`

Optional preset template id. Currently supports 'startup-speed'.

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: `operations`, `orchestration`

***

## General

### orgx\_agent\_sessions

List active CLI session IDs stored for workstreams. Used for session resume support.

**initiativeId** `string`

Optional initiative ID filter.

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: `operations`, `orchestration`

***

### orgx\_clear\_agent\_session

Clear stored CLI session IDs for an initiative (forces fresh sessions on next dispatch).

**initiativeId** `string` *required*

Initiative UUID

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: `orchestration`

***

### orgx\_get\_morning\_brief

Get the morning brief value surface: session summary, value signals, exceptions, trust events, and top receipts. Preferred replacement for ROI summary lookups.

**workspace\_id** `string` *required*

Workspace ID to load the brief for

**session\_id** `string`

Optional autonomous session ID

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_list\_entities

List OrgX entities of a given type with optional status filter.

**type** `string` *required*

Entity type: initiative, workstream, task, decision, agent

**status** `string`

Filter by status

**limit** `number`

Max results (default 20) Default: `20`.

**initiative\_id** `string`

Filter by initiative ID

**project\_id** `string`

Legacy project/workspace scope filter

**workspace\_id** `string`

Workspace ID (canonical scope)

**command\_center\_id** `string`

Deprecated alias for workspace\_id

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: Not exposed on scoped domain endpoints

***

### orgx\_query\_org\_memory

Search OrgX organizational memory (decisions, initiatives, artifacts). Preferred replacement for older decision-history lookups.

**query** `string` *required*

What to search for

**scope** `string`

Optional memory scope Options: `all`, `artifacts`, `decisions`, `initiatives`.

**limit** `number`

Max results to return

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_recommend\_next\_action

Get the preferred next action for a workspace, initiative, workstream, or milestone. Preferred replacement for raw queue scoring in operator workflows.

**entity\_type** `string`

Entity type to recommend for (default: workspace) Options: `workspace`,
`initiative`, `workstream`, `milestone`.

**entity\_id** `string`

Entity ID. For workspace, use "default" or a workspace ID.

**workspace\_id** `string`

Optional canonical workspace scope

**command\_center\_id** `string`

Deprecated alias for workspace\_id

**agent\_id** `string`

Optional canonical agent ID used to filter recommendations

**domain** `string`

Optional agent domain used to filter recommendations Options: `engineering`,
`product`, `design`, `marketing`, `sales`, `operations`, `orchestration`.

**canonical\_only** `boolean`

Return only the next assigned task for the active goal

**limit** `number`

Max recommendations to return

**cascade** `boolean`

Refresh recommendations across the entity chain first

**Scope Class**: Base domain tool (available in every domain-scoped endpoint)

**Scoped Domains**: `design`, `engineering`, `marketing`, `operations`, `orchestration`, `product`, `sales`

***

### orgx\_resume\_agent\_session

Store or update a CLI session ID for a workstream so the next slice resumes it.

**workstreamId** `string` *required*

Workstream UUID

**sessionId** `string` *required*

CLI session UUID to resume

**initiativeId** `string` *required*

Initiative UUID

**sourceClient** `string`

Source client (codex, claude-code)

**Scope Class**: Domain-specific or unscoped-only

**Scoped Domains**: `orchestration`

***


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.