How do I use sop-mcp?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@sop-mcp run the code review process" 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.

sop-mcp

by ValueArchitectsAI

Overview Schema Related Servers Score Discussions

Python

Local

sop-mcp

PyPI Python License

An MCP server that brings process automation to AI agents through Standard Operating Procedures.

LLMs are powerful but unpredictable when executing multi-step processes — they skip steps, summarize instead of act, and lose track of where they are. sop-mcp solves this by delivering procedures one step at a time, forcing the agent to execute each step and provide concrete output before advancing. This turns SOPs into a control mechanism that makes LLM behavior predictable and auditable.

The result: agents that follow processes the way humans do — step by step, with reasoning enforced at each level.

This approach aligns with Agent SOPs — a standardized markdown format for defining AI agent workflows using RFC 2119 requirement levels (MUST, SHOULD, MAY). sop-mcp adds the execution layer: an MCP server that delivers these procedures one step at a time and enforces completion before advancing.

Install

Kiro	Cursor	VS Code

Or add manually:

{
  "mcpServers": {
    "sop-mcp": {
      "command": "uvx",
      "args": ["sop-mcp"],
      "env": { "SOP_STORAGE_DIR": "/path/to/your/sops" }
    }
  }
}

Related MCP server: Stratum MCP Server

How It Works

Every session starts the same way — discover what's available, then execute.

list_resources()                                     → catalog of sop:// URIs
run_sop(sop_name="sop_creation_guide")               → Step 1 + instructions
run_sop(..., current_step=1, step_output="...")      → Step 2
run_sop(..., current_step=2, step_output="...")      → Step 3
  ...
run_sop(..., current_step=N, step_output="...")      → Completion

Each response tells the agent to execute the step — not just read it.

Bundled SOPs

Four SOPs ship with the server so new users can try run_sop immediately:

SOP	What it does
`sop_creation_guide`	Step-by-step guide for authoring new SOPs with RFC 2119 requirements
`code_review_process`	Standard code review workflow — prepare, review, address feedback, merge
`employee_onboarding_setup`	IT setup for a new hire — alias, email, hardware selection
`user_onboarding_process`	Provision identity, application access, and welcome package

Storage default: ~/.sop_mcp (seeded from the bundled SOPs on first run). Override with SOP_STORAGE_DIR.

Tools

Tool	Purpose
`list_resources`	Discover available SOPs (built in to every MCP client)
`read_resource`	Read an SOP's full content before executing it
`run_sop`	Execute an SOP step by step
`publish_sop`	Create or update an SOP
`submit_sop_feedback`	Record improvement suggestions

Full parameter reference: docs/mcp-reference.md

Documentation

Audience	Resource
AI tools	`llms.txt` — auto-discovered server description
Users	`skills/sop-mcp-usage/` — how to use
Operators	`skills/sop-mcp-configuration/` — install, configure, hooks
Developers	`CONTRIBUTING.md` — build, test, design decisions
Reference	`docs/mcp-reference.md` — full tool schemas

Development

uv sync              # install dependencies
uv run pytest        # run tests
uv run sop-mcp       # start server locally
uv run python scripts/generate_docs.py  # regenerate docs

License

MIT

Available Tools

5 tools

list_resourcesA

List all available resources with their URIs and descriptions.

ParametersJSON Schema

Name	Required	Description	Default
No parameters

TDQS

A3.8/5.0

Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the full burden of behavioral disclosure. It states that the tool lists resources and returns URIs and descriptions, but does not mention whether there is pagination, rate limiting, or any side effects. As a read-only listing, this is somewhat implicit but not fully transparent.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, front-loaded sentence that directly states the purpose and output without any unnecessary words or redundancy. It is perfectly concise for the simplicity of the tool.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's simplicity (no parameters, no output schema), the description adequately states what it does and what it returns. It could be slightly more complete by clarifying what constitutes a 'resource' in this context, but it is sufficient for an agent to invoke it correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The tool has zero parameters and an empty input schema, so there are no parameter semantics to clarify. Per the baseline for zero parameters, a score of 4 is appropriate; the description doesn't need to explain any inputs.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description specifies the action as 'list' with a clear resource type ('all available resources') and the included details (URIs and descriptions). It is distinct from sibling tools like read_resource, which implies a targeted access, while this tool provides an overview.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives no explicit guidance on when to use this tool versus alternatives. It does not mention that this is for discovery before using read_resource, nor does it state any exclusions or conditions. The usage context is only implied by the word 'list'.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

publish_sopA

Publish a new or updated Standard Operating Procedure document.

The content parameter MUST contain the complete SOP markdown string with YAML frontmatter declaring:

name (required, snake_case, ≥3 underscore segments — the SOP's identity)
owner (required, non-empty string — team, alias, or email. This is the point of contact surfaced when feedback is submitted or a mismatch is detected during review. Pick a name you want pinged.)
stage (required, 'preprod' or 'prod' — informational lifecycle label; see the stage argument below for mismatch behaviour)
version (required, positive integer — advisory revision counter. The tool auto-bumps on each publish (+1), but we ask authors to declare it explicitly so a mismatch between the file on disk and what the author thinks they are updating is visible in the response)
description (optional — when omitted, the SOP's ## Overview section is used for short summaries)

Version & stage mismatch: the tool never trusts the frontmatter values blindly. The stage argument wins over the frontmatter stage, and the version is computed server-side (max existing + 1). Both values are overwritten in the stored content so the file on disk always reflects what actually happened. If you pass a version or stage that disagrees with the final stored values, the response surfaces the difference under warning so you can decide whether you were editing the right version.

Example call: {"content": "---\nname: my_sop_name\nversion: 1\nowner: my-team\nstage: preprod\n---\n\n# My SOP\n\n## Overview\nOverview text.\n\n### Step 1: First step\nDo the thing."}

Versioning: plain positive integers — 1, 2, 3, 4, … New SOPs start at 1; each subsequent publish increments by one. No semver.

Lint enforcement: every publish runs the same rule engine as the standalone sop-lint CLI. Errors (SOP rules at severity=error) BLOCK the publish — the tool raises and nothing is written. Warnings are returned under the warning field but do not block. Iterate locally with sop-lint <file> before calling publish_sop to avoid MCP round-trip latency.

ParametersJSON Schema

Name	Required	Description	Default
`stage`	Yes	Deployment stage: 'preprod' or 'prod'
`content`	Yes	Complete SOP markdown with YAML frontmatter (name, owner, stage, version)

TDQS

A4.8/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full burden. It thoroughly discloses server-side version computation, stage argument precedence, content overwriting, lint errors blocking publish, and warning field behavior. This is exceptional transparency for a mutation tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is front-loaded with a one-sentence purpose, then uses bold-labeled sections to organize complex details. Every sentence earns its place, including the example and lint guidance. Despite length, it is well-structured and not padded.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given that there is no output schema and no annotations, the description covers all critical aspects: required input format, versioning and stage precedence, error/warning behavior, and local linting workflow. It is practically complete for correct invocation and outcome understanding.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The schema descriptions are minimal, but the description adds extensive semantics: required YAML frontmatter fields with types and rules, versioning behavior, stage mismatch handling, and a concrete example call. This far exceeds the schema's surface-level descriptions.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The first sentence clearly states the action ('Publish') and the resource ('a new or updated Standard Operating Procedure document'). It distinguishes from sibling tools like submit_sop_feedback and read_resource, making the tool's role unambiguous.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives clear context: prepare content with required frontmatter and iterate with sop-lint locally before calling to avoid latency. However, it does not explicitly compare with sibling alternatives or state when not to use this tool, so it lacks explicit exclusions.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

read_resourceC

Read a resource by its URI.

ParametersJSON Schema

Name	Required	Description	Default
`uri`	Yes

TDQS

C2.7/5.0

Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description must carry the full burden of behavioral disclosure. It only says 'Read', implying a non-destructive operation, but provides no details about error behavior, authentication requirements, or resource existence handling. The description adds minimal value beyond what the name already suggests.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single sentence with no wasted words, achieving high conciseness. However, it is so brief that it borders on under-specification. For the purpose of being concise, it scores well, though it lacks substance.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the simple schema and lack of output schema or annotations, the description is quite incomplete. It does not explain what happens when a resource is not found, whether the URI must be fully qualified, or any other context. This makes the tool minimally viable but with clear gaps.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters2/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The schema has 0% description coverage and the parameter is just a required 'uri' string. The description mentions 'by its URI', which is minimally helpful but does not explain what constitutes a valid URI, any formats, or examples. It adds little beyond the schema's parameter name and type.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description uses a specific verb ('Read') and resource ('resource'), and clarifies the identifier type ('URI'). While it doesn't explicitly differentiate from sibling tools like list_resources, the name and action imply a single-resource read operation, which is reasonably clear.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives no guidance on when to use this tool versus alternatives. It merely states what it does without any context about appropriate use cases, limitations, or exclusions. Sibling tools like list_resources could overlap, but no differentiation or selection criteria are provided.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

run_sopA

Start or advance a Standard Operating Procedure step by step. Use list_resources to discover available SOPs, then call this tool with the SOP name.

Each call returns one step. Execute the step, then call again with current_step incremented to advance.

IMPORTANT: You MUST execute ALL actions described in the returned step content. Do NOT just read or summarize the step — perform the actions using your available tools.

When continuing (current_step >= 1), you MUST provide step_output with the concrete output you produced for the completed step.

ParametersJSON Schema

Name	Required	Description
`sop_name`	Yes	Name of the SOP to execute (use list_resources to discover available SOPs)
`step_output`	No	Concrete output you produced for the completed step. Required when current_step >= 1
`current_step`	No	Step number to advance from. 0 to start, N to advance past step N

TDQS

A4.4/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the full burden of behavioral disclosure. It transparently explains that each call returns one step, that current_step must be incremented to advance, and that all actions in the step must be executed (not just read). It also mandates supplying step_output when continuing. This is rich, actionable behavioral context beyond the schema.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is well-structured and front-loaded with the primary purpose. It uses short, direct sentences and scannable formatting (IMPORTANT warning) to highlight critical execution requirements. Every sentence provides necessary operational guidance without redundancy or fluff.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The description covers the core usage loop, prerequisites, and required outputs, which is strong for a stateful tool without an output schema. However, it does not explain how to determine when the SOP is complete or what the returned step content looks like structurally (e.g., text vs. structured actions). This is a minor gap but not severe given the detailed procedural guidance.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The schema description coverage is 100%, so the parameters are already well-documented. The description adds no new meaning beyond what is in the schema, merely restating the same information about current_step and step_output. It reinforces the stateful usage but does not deepen parameter understanding, so the baseline of 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's function: 'Start or advance a Standard Operating Procedure step by step.' It also specifies the resource (SOP) and the step-wise execution model. The usage of 'discover available SOPs' via list_resources helps distinguish it from sibling tools like read_resource or publish_sop.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It explicitly instructs to use list_resources to discover SOPs before calling this tool, providing a clear prerequisite. It also explains the step-by-step advancement process and when step_output is required. However, it does not explicitly contrast with alternatives like read_resource or submit_sop_feedback, though the workflow is strong enough to imply appropriate usage.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

submit_sop_feedbackA

Submit improvement feedback for a specific SOP.

Feedback is appended as a single JSON line to {sop_name}.feedback.jsonl inside the SOP's folder. Each entry captures the SOP version, a UTC timestamp, and the feedback text — ready for review when the SOP is next revised.

ParametersJSON Schema

Name	Required	Description	Default
`feedback`	Yes	Improvement feedback text — what worked, what needs fixing
`sop_name`	Yes	Name of the SOP to submit feedback for

TDQS

A4/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Despite having no annotations, the description precisely discloses behavioral details: feedback is appended as a JSON line to a .feedback.jsonl file in the SOP's folder, including version, timestamp, and text. This goes beyond a generic statement and gives clear expectations for the side effect.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is two sentences that front-load the purpose and then add an efficient behavioral detail. No wasted words; every clause adds necessary context.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's simplicity and full schema coverage, the description covers purpose, mechanics, and side effects. It could optionally mention validation or error conditions, but these are not required for a basic feedback submission tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Both parameters are fully documented in the schema with clear descriptions (sop_name and feedback). The description does not add significant additional parameter-level semantics beyond what the schema already provides, so it remains at the schema-corrected baseline.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description uses a specific verb 'Submit' with a resource 'improvement feedback for a specific SOP', clearly differentiating it from sibling tools like list_resources, run_sop, and publish_sop. It states exactly what the tool does without ambiguity.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies usage when one has improvement feedback for an SOP but does not explicitly mention alternatives or when-not-to-use scenarios. It provides context that feedback is for future review, but lacks a direct comparison to sibling tools.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.

5 tool updatesv0.12.2Aug 12, 2026
- First observedlist_resources
- First observedpublish_sop
- First observedread_resource
- First observedrun_sop
- First observedsubmit_sop_feedback

Install Server

RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

A3.8/5.0

Disambiguation5/5

Each tool targets a distinct action: publishing, running, providing feedback, and generic resource listing/reading. No overlaps or ambiguous boundaries exist between the tools.

Naming Consistency5/5

All tools use verb_noun snake_case, with run_sop, publish_sop, list_resources, read_resource following the pattern, and submit_sop_feedback as a natural extension. Consistent and predictable.

Tool Count5/5

Five tools cover the SOP lifecycle without bloat; each tool serves a clear purpose. The count is well within the ideal range for a focused server.

Completeness4/5

The surface covers publish (create/update), run (execution), and feedback, with list/read for discovery. Missing delete/archive and feedback retrieval, but those are minor gaps for the core workflow.

Maintenance

ActivityActive

ResponsivenessNo issues

Resources

GitHub Repository

Need Help?

Related Servers

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

AgentFabric — Human-in-the-Loop Review & Approval B
io.github.enzoduit
Human-in-the-loop review and approval for AI agents. Audit trail, approval policies, native MCP.
Agent Orchestration Security Observability
11
governance-platform A
ai.geodesiclabs
Pre-execution governance for AI agents. Deterministic PASS/FAIL/REVIEW verdicts, replayable proof.
AI & Machine Learning Testing & QA Tools Monitoring
373
Warden
io.github.Larshiensch99
Register every AI agent, log every action, prove it. EU AI Act compliance built in.
Observability Security
2
DobbyOAuth
io.github.gil-dobby
Connect, monitor, and control AI agents — tasks, approvals, schedules, and governance.
Agent Orchestration Autonomous Agents Monitoring
2

Related MCP Servers

promptspeak-mcp-server
Autonomous Agents Security App Automation
chrbailey
A
license
B
quality
C
maintenance
Pre-execution governance for AI agents. 45 MCP tools for hold queues, audit trails, risk scoring, and policy enforcement. Validates agent actions before they execute.
Updated 2 months ago (2026-07-05 20:38 UTC)
45
118
1
MIT
Stratum MCP Server
Coding Agents Autonomous Agents
smartmemory
A
license
Not graded
quality
B
maintenance
Enables AI coding agents to execute formal, stateful workflows with typed contracts, postcondition enforcement, and structured retry logic.
Updated 2 months ago (2026-07-12 11:10 UTC)
1
Apache 2.0
Workflow MCP Server
App Automation Autonomous Agents
tuanknguyen
A
license
B
quality
D
maintenance
Guides agents through structured workflows with predefined steps, allowing for flexible execution via instructions and tool calls.
Updated a year ago (2025-04-08 04:30 UTC)
8
1
MIT
Workflow Orchestration MCP
Agent Orchestration Developer Tools
m2ux
A
license
Not graded
quality
B
maintenance
Guides AI agents through structured, multi-step workflows with discovery, navigation, and fidelity enforcement.
Updated 15 hours ago (2026-09-05 16:14 UTC)
1
Apache 2.0

Latest Blog Posts

Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)
By Om-Shree-0709 on July 25, 2026.
mcp
Agent Identity
OAuth 2.1
Your AI Chatbot Just Exposed Your CEO's Salary to an Intern
By Om-Shree-0709 on July 2, 2026.
Agent Identity
MCP Security
OAuth Delegation
Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)
By Om-Shree-0709 on June 30, 2026.
Agentic Ai
Prompt Injection
WebAssembly

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/ValueArchitectsAI/sop-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server