Provides governance for CrewAI agents, wrapping agent actions with pre-execution gates, audit traces, and rollback capabilities to ensure compliance with regulations.
Provides governance for LangGraph agents, integrating pre-execution policy checks, audit trails, and rollback support for regulated workflows.
Any agent. Any framework. Governed by default.
Every AI agent action — writing files, calling APIs, sending emails, modifying data — executes without a receipt. You don't know what ran, what was blocked, or why.
DingDawg Governance adds a pre-execution gate that:
Blocks policy violations before they execute — fail-closed, not fail-open
Generates LNN causal traces — interpretable reasoning chain for every decision
Issues IPFS audit proofs — tamper-evident receipts pinned to distributed storage
Supports rollback — every governed action carries enough context to reverse it
Assigns @handle identities — agents get a governed identity (@billing-agent, @hr-screener) with a full action history tied to that handle
Related MCP server: dingdawg-loop
Built for frameworks where AI agent decisions carry legal weight:
Industry | Regulation |
Healthcare | HIPAA — PHI access, treatment decision logging |
Insurance / Fintech | State regulations, adverse action documentation |
Employment | CO SB 205, EEOC — automated hiring decision audit |
Legal | Chain-of-custody, privileged data access controls |
Edtech | FERPA — student data access receipts |
npm install dingdawg-governancepip install dingdawg-loopAdd to ~/.claude/mcp.json or project-level .mcp.json:
{
"mcpServers": {
"dingdawg-governance": {
"command": "npx",
"args": ["dingdawg-governance"],
"env": {
"DINGDAWG_API_KEY": "your-api-key"
}
}
}
}Without an API key, all tools work locally. Receipts stored at ~/.dingdawg/governance/receipts/.
from dingdawg_loop import schedule_governed
@schedule_governed(
agent_id="@data-sync-agent",
cron="0 * * * *",
risk_tier="medium"
)
def sync_records():
# Your agent logic here
passTwo lines. Every execution is pre-checked, receipted, and fail-closed. If governance denies, the function does not run.
Tool | What it does |
| Pre-execution gate — evaluates risk, issues receipt, blocks on violation |
| Retrieve receipts by agent handle, time range, or receipt ID |
| Score against EU AI Act, CO SB 205, NIST AI RMF, ISO 42001 |
| Reverse a governed action using its receipt context |
| Assign a governed @handle identity to an agent |
| Retrieve or pin IPFS audit proof for a receipt |
Layer | License | Where |
SDK core (govern, audit, compliance) | Apache 2.0 | This repo |
LNN causal trace engine | Cloud only | |
IPFS proof pinning | Cloud only | |
Team audit trail + cross-agent history | Cloud only | |
Compliance report PDFs (certified) | Paid tier |
The core gate runs fully offline. Cloud unlocks team visibility, IPFS pinning, and certified compliance reports.
Runnable examples in examples/:
File | What it shows | Regulated use case |
| Fintech — payment transfer gate | |
| Healthcare — HIPAA PHI sync | |
CrewAI agents wrapped with governance | Employment — CO SB 205 hiring audit | |
Drop-in | All regulated verticals |
Each example includes expected output as comments and the governance receipt structure.
dingdawg.com — platform, pricing, API keys
dingdawg.com/docs/integrations — CrewAI, LangGraph, Cursor, Claude Code
dingdawg.com/harness — LNN engine, IPFS proofs, advanced governance
dingdawg.com/compliance — CO SB 205 gap report ($199)
Get the governance audit trail. Returns governance receipts from local storage or cloud API. Free to use.
| Name | Required | Description | Default |
|---|---|---|---|
| receipt_id | No | Receipt ID from a govern_action call | |
| agent_id | No | Agent ID to get all governed actions for | |
| time_range | No | Time range for audit trail lookup | |
| limit | No | Maximum number of records to return (default 10) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description must fully disclose behavioral traits. It states the data source (local storage or cloud API) and that it is free, but fails to mention whether the operation is read-only, any authentication requirements, rate limits, or side effects. The description lacks important 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 extremely concise with two sentences: the first states the main purpose, the second adds a key feature (returns from two sources) and a cost note. Every sentence adds value without redundancy. It is well front-loaded.
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 four parameters, no output schema, and no annotations, the description lacks completeness. It does not describe the return value format, pagination behavior, or how parameters interact (e.g., using receipt_id vs agent_id). More detail is needed for an agent to fully understand the tool's behavior.
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 input schema covers all four parameters with descriptions, meeting 100% schema description coverage. The tool description does not add significant extra meaning beyond the schema; it provides contextual background about governance receipts and storage sources but does not elaborate on parameter usage or constraints.
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 verb 'Get' and the resource 'governance audit trail', providing a specific action and object. It distinguishes from sibling tools by focusing on retrieving receipts, whereas siblings like 'govern_action' and 'compliance_check' involve different operations.
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 mentions 'Free to use' regarding cost, but does not provide any guidance on when to use this tool over its siblings (compliance_check, govern_action) or under what circumstances it is appropriate. No exclusions or prerequisites are stated.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Quick compliance check against common AI governance frameworks. Free tier: 10 checks per day. Evaluates against EU AI Act, Colorado AI Act, NIST AI RMF, and ISO 42001.
| Name | Required | Description | Default |
|---|---|---|---|
| system_description | Yes | Describe your AI system: what it does, data sources, decision scope | |
| framework | No | Framework to check against (default: all) | |
| deployment_stage | No | Current deployment stage |
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 discloses a rate limit (10 checks per day) but does not explicitly state whether the tool is read-only or destructive. It also doesn't describe the output format or side effects, leaving some ambiguity for a compliance check 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?
The description is two sentences long, with no unnecessary words. The first sentence states the purpose, and the second adds constraints and scope. It is front-loaded and efficient.
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 no output schema and no annotations, the description is fairly complete for a simple check tool. However, it lacks details about the return format (e.g., pass/fail, risk score) and any prerequisites for using the tool. It touches on constraints but not on usage context beyond the free tier.
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?
All three parameters are fully described in the input schema (100% coverage). The description adds no additional meaning beyond what the schema already provides, such as the framework enum and descriptions. Therefore, it meets the baseline for schema coverage.
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 it performs a 'quick compliance check against common AI governance frameworks', specifying the exact frameworks (EU AI Act, Colorado AI Act, NIST AI RMF, ISO 42001). The verb 'check' distinguishes it from sibling tools (audit_trail, govern_action) which imply deeper audit or action.
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 does not explicitly say when to use this tool versus siblings. It implies a quick preliminary check but provides no guidance on when to prefer it over audit_trail or govern_action. The free tier limit is mentioned, which indirectly suggests usage constraints.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Govern any AI agent action. Performs capability check + policy evaluation + generates a governance receipt. Returns a receipt proving the action was governed. When API key is set, uses cloud API with local fallback.
| Name | Required | Description | Default |
|---|---|---|---|
| agent_id | Yes | Identifier for the AI agent performing the action | |
| action_type | Yes | Type of action (e.g., 'send_email', 'make_purchase', 'modify_data', 'api_call') | |
| action_description | Yes | Human-readable description of what the agent is about to do | |
| target_resource | No | The resource being acted upon (e.g., 'user_database', 'email_server', 'payment_api') | |
| risk_tier | No | Self-assessed risk level of this action | |
| context | No | Additional context key-value pairs for policy evaluation |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Discloses key behaviors: capability check, policy evaluation, receipt generation, and cloud fallback when API key is set. No annotations exist, so description carries the full burden; it adequately informs about functionality without side-effect warnings.
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?
Two sentences, front-loaded with primary purpose followed by key behavioral detail (fallback mechanism). Every sentence adds value, 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?
Explains the return value (receipt) and process. Lacks detail on error conditions or policy failure handling. No output schema, but description suffices for basic understanding.
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 100%, so baseline is 3. The description adds no extra parameter details beyond what the schema provides, earning no bonus.
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?
Description clearly states the tool governs any AI agent action, performing capability checks, policy evaluation, and generating a governance receipt. It distinguishes from sibling tools (audit_trail, compliance_check) by focusing on proactive governance with a receipt output.
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?
Implies the tool is used for governing AI actions but does not explicitly contrast with siblings or provide when-not-to-use guidance. No alternative tools are suggested for different scenarios.
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. Dates show when Glama detected each change.
v1.0.3audit_trailcompliance_checkgovern_actionEach tool serves a distinct function: audit_trail for retrieving records, compliance_check for framework evaluation, and govern_action for governing an action. No overlap in purposes.
All tool names follow a consistent verb_noun pattern with snake_case: audit_trail, compliance_check, govern_action. Predictable and uniform.
Three tools is ideal for a focused governance server, covering auditing, compliance checks, and action governance without redundancy or bloat.
The set covers core governance workflows (audit, compliance, action governance). Minor gaps exist (e.g., no tool to manage or delete receipts), but the surface is coherent and functional for its scope.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Pre-execution governance for AI agents. Deterministic PASS/FAIL/REVIEW verdicts, replayable proof.
Agent governance with A2A/Shopify/MCP trust audits, action screening, and decision UI.
Sovereign Agent OS — Persistent Memory, Governance & Compliance for AI Agents.
Blockchain SSN for AI agents. MCP gateway that blocks at the point of action, tamper evident audit.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/dingdawg/governance-sdk'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
