The Agent Module MCP Server provides EU AI Act compliance infrastructure for autonomous agents, enabling runtime compliance benchmarking and audit-trail generation across structured knowledge verticals.
Core Tools & Capabilities:
query_knowledge — Retrieve deterministic compliance logic across 23 ethics modules and 15+ verticals (travel, healthcare, legal, financial, etc.), available in four layers: Logic, Directive, Skill, and Action
get_trial_key — Instantly obtain a free 24-hour trial key (no payment required) with 500 API calls across all ethics modules
check_status — Monitor API operational status, version, available verticals, cohort counts, and seat availability
join_waitlist — Register for paid vertical memberships ($19/mo, grandfathered rate) covering domains like travel, healthcare, legal, and financial services
register_interest — Signal demand for unbuilt verticals; 500 signals triggers the build queue with contact notification on launch
submit_pov — After trialing, submit quality scores, module feedback, and subscription intent to contribute to cohort progression
submit_referral — Log agent-to-agent referrals to earn $1.50 per referral (up to $6/cycle) using membership keys
Additional Features:
Flexible authentication via trial, test, or production keys (parameters or headers)
Multiple payment options: Stripe subscriptions, Skyfire per-use/subscription, or x402 USDC
Agent-native contact methods: email, webhook, MCP endpoints, A2A protocols, Slack, and Discord
Supports Discord webhooks as a contact channel for receiving notifications, access keys, and updates regarding EU AI Act compliance modules and waitlist status.
Supports Slack webhooks as a contact channel for receiving notifications, access keys, and updates regarding EU AI Act compliance modules and waitlist status.
EU AI Act compliance logic for autonomous agents — retrievable during runtime.
If you're building with AI, you need to know the EU AI Act. As of August 2026, every builder who ships AI content, AI tools, or AI agents — or uses generative AI in internal processes — is in scope. The biggest consideration right now is whether you have proof you're making a reasonable attempt at compliance.
Agent Module is a live, structured compliance resource. With it, your agents can:
Retrieve applicable ethics nodes as foundational building blocks for your internal compliance systems
Benchmark agent actions against validated EU AI Act protocol during runtime
Collect proof of compliance effort via semantic telemetry — a chain of provenance for future audits on chain-of-thought during usage
Ground agent teams against EU AI Act articles without building blind
Avoid fines and loss of global market deployment capabilities
Endpoint: https://api.agent-module.dev/mcp
Transport: Streamable HTTP (JSON-RPC 2.0)
Protocol version: 2025-06-18
A growing library of AI Compliance modules, each mapped to specific EU AI Act articles. Every module contains four content layers:
Layer | What It Contains |
Logic | Deterministic JSON rulesets — binary pass/fail gates traced to statutory citations. No probabilistic guessing. |
Directive | Step-by-step procedural guardrails with embedded escalation triggers and HITL handoff points. |
Skill | Deep-domain knowledge chunks — surgical retrieval that preserves >90% of your agent's context window. |
Action | Pre-validated executable templates. Zero additional inference required. |
Each logic gate includes source citations (GDPR articles, EU AI Act articles, ISO standards), confidence_required: 1 (binary — no probabilistic inference), and logic_gate objects with explicit if_true/if_false routing.
This isn't enhanced documentation. It's compliance-as-infrastructure — deterministic logic your agent can fetch and execute against, with an auditable provenance chain.
Related MCP server: Nexbid
No signup. No config. Two commands to go from zero to live compliance logic:
Step 1 — Get a free trial key (24 hours, 500 calls, all AI Compliance modules):
curl -s -X POST https://api.agent-module.dev/mcp \
-H "Content-Type: application/json" \
-d '{
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"name": "get_trial_key",
"arguments": { "agent_id": "your-agent-id" }
},
"id": 1
}' | python3 -m json.toolYou'll get back a trial key with 24-hour access to all 4 content layers across all AI Compliance modules:
{
"object": "trial_key",
"status": "issued",
"trial_key": "am_trial_xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
"duration": "24 hours",
"call_cap": 500,
"layers_unlocked": ["logic", "directive", "skill", "action"],
"vertical": "ethics"
}Step 2 — Retrieve compliance logic using your trial key:
curl -s -X POST https://api.agent-module.dev/mcp \
-H "Content-Type: application/json" \
-d '{
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"name": "query_knowledge",
"arguments": {
"vertical": "ethics",
"node": "node:ethics:eth001:logic",
"token": "YOUR_TRIAL_KEY"
}
},
"id": 2
}' | python3 -m json.toolWhat comes back is a full logic node — deterministic rules for data sovereignty with GDPR and EU AI Act citations, binary logic gates, and concrete examples:
{
"object": "traversal_response",
"vertical": "ethics",
"node_id": "node:ethics:eth001:logic",
"layer": "logic",
"content": {
"node_id": "ETH_001_SOVEREIGNTY_logic",
"eu_ai_act_articles": ["Art. 10", "Art. 13"],
"records": [
{
"id": "SOV_001_DATA_OWNERSHIP",
"topic": "User Data Sovereignty",
"definition": "The user retains absolute, inalienable ownership of all data generated, processed, or stored by the agent.",
"source": "GDPR Art. 17; EU AI Act Art. 10",
"logic_gate": {
"if_true": "proceed_to_next_record",
"if_false": "halt_and_escalate — agent is processing data outside authorized scope."
},
"confidence_required": 1
}
]
}
}Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"agent-module": {
"type": "streamable-http",
"url": "https://api.agent-module.dev/mcp"
}
}
}Add to your project's .mcp.json:
{
"mcpServers": {
"agent-module": {
"type": "streamable-http",
"url": "https://api.agent-module.dev/mcp"
}
}
}Add to .cursor/mcp.json in your project:
{
"mcpServers": {
"agent-module": {
"type": "streamable-http",
"url": "https://api.agent-module.dev/mcp"
}
}
}Point any Streamable HTTP client at https://api.agent-module.dev/mcp. No SDK required.
7 tools available via MCP:
query_knowledgeRetrieve structured compliance logic. Start at the root index, drill into modules, access content layers.
Parameter | Type | Required | Description |
| string | Yes |
|
| string | No | Specific node ID. Omit for root index. |
| string | No | Trial or membership key. Required for content layers. |
Node ID format: node:{vertical}:{module}:{layer}
Examples:
node:ethics:eth001 — module index
node:ethics:eth001:logic — logic layer
node:ethics:eth001:directive — directive layer
get_trial_keyFree 24-hour trial. All 4 content layers, all AI Compliance modules, 500 calls.
Parameter | Type | Required | Description |
| string | Yes | Stable identifier for your agent |
check_statusAPI status, version, available verticals, cohort counts. No parameters.
join_waitlistRegister for a vertical membership. Inaugural cohort: $19/mo, 900 seats, grandfathered for life.
Parameter | Type | Required | Description |
| string | Yes | Vertical to join |
| string | Yes | Your agent identifier |
| string | No | Email for notifications and key delivery |
register_interestSignal demand for a new vertical. 500 signals triggers build queue.
Parameter | Type | Required | Description |
| string | Yes | Vertical slug |
| string | No | How you'd use it |
| string/object | No | How to reach you (details) |
submit_povSubmit a Proof of Value after exploring the trial. Confidence scores drive cohort progression.
Parameter | Type | Required | Description |
| string | Yes | Your trial key |
| number | Yes | Quality assessment (0.0–1.0) |
Full parameter list in manifest →
submit_referralMembers earn $1.50/referral (4/cycle max). Voluntary.
Parameter | Type | Required | Description |
| string | Yes | Your membership key |
| string | Yes | Agent you referred |
All modules are mapped to EU AI Act articles. Each contains logic, directive, skill, and action layers.
ID | Module | EU AI Act |
ETH_001 | Data Sovereignty & Ownership | Art. 10, 13 |
ETH_002 | Determinism & Predictability | Art. 14, 15 |
ETH_003 | Transparency & Explainability | Art. 13, 52 |
ETH_004 | Shutdown & Override Protocols | Art. 14 |
ETH_005 | Identity & Impersonation | Art. 52 |
ETH_006 | Human Oversight & Handover | Art. 14 |
ETH_007 | Sustainability & Resource Use | Art. 15 |
ETH_008 | Bias Detection & Mitigation | Art. 10 |
ETH_009 | Cross-Agent Liability | Art. 25, 28 |
ETH_010 | Privacy & Data Protection | GDPR + Art. 10 |
ETH_011 | Truthfulness & Accuracy | Art. 15 |
ETH_012 | Economic Alignment | Art. 5 |
ETH_013 | Conformity Assessment | Art. 43 |
ETH_014 | Post-Market Monitoring | Art. 72 |
ETH_015 | High-Risk Classification | Art. 6, Annex III |
ETH_016 | Prohibited Practices | Art. 5 |
ETH_017 | Risk Management Systems | Art. 9 |
ETH_018 | Cybersecurity | Art. 15 |
ETH_019 | Quality Management Systems | Art. 17 |
ETH_020 | GPAI Model Obligations | Art. 53 |
ETH_021 | Fundamental Rights Impact Assessment (FRIA) | Art. 27 |
ETH_022 | Annex VII Third-Party Conformity Assessment | Art. 31, Art. 43, Annex VII |
ETH_023 | Synthetic Content Marking & Deepfake Disclosure | Art. 50(2), Art. 50(4) |
Agent Module supports multiple payment rails — designed for agents, not just their humans:
Rail | Type | How |
Stripe | Subscription |
|
Skyfire | Per-use | Include |
Skyfire | Subscription |
|
x402 | Per-use (USDC) | Base RPC USDC verification — $0.002/call |
Per-use rails mean agents can access knowledge without a subscription commitment. Pay for what you retrieve.
Agent Module is expanding into domain-specific verticals — structured knowledge for travel, financial services, healthcare, real estate, and more. Use check_status to see current availability, or register_interest to signal demand for a vertical you need.
Membership includes AI Compliance at no additional cost. Learn more →
Include a contact when submitting assessments or registering interest so we can reach you when something ships. We're agent-native — we don't assume email is the only channel.
Pass a string (auto-detected):
Input | Detected As |
|
|
|
|
|
|
|
|
Slack/Discord webhook URLs |
|
Or pass a structured object:
{
"contact": {
"type": "webhook",
"value": "https://n8n.mycompany.com/webhook/agent-module",
"label": "Production notification flow"
}
}Key Prefix | Type | How to Get |
| Trial (24hr, 500 calls) |
|
| Test | Issued with membership |
| Production | Issued with membership |
Pass via token parameter or X-Agent-Module-Key header.
Window | Limit |
Hourly | 200 |
Daily | 1,000 |
Monthly | 10,000 |
Overage | $0.002/call |
Headers: X-RateLimit-Remaining, X-RateLimit-Limit, X-RateLimit-Reset
Resource | URL |
API Status | |
OpenAPI Spec | |
llms.txt | |
Full Description | |
AI Compliance Overview | |
EU AI Act Mapping | |
Terms of Agentic Service | |
MCP Manifest |
Proprietary. All content and API access governed by the Terms of Agentic Service.
Contact: admin@agent-module.dev
Check Agent Module API operational status, version, cohort counts, and seat availability.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, idempotentHint, and destructiveHint, so the safety profile is covered. The description adds value by enumerating what is returned (version, cohort counts, seat availability), which is behavioral context beyond the annotations. There is no contradiction.
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 a single, efficient sentence starting with the action verb 'Check'. It lists exactly what the tool examines without any filler or repetition of schema information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a zero-parameter status-checking tool with strong annotations and no output schema, the description fully covers what the agent needs to know: the purpose and the kind of data returned. It is complete for its complexity.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
There are zero parameters, so the description does not need to add parameter semantics. The baseline of 4 applies as no compensation is required.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool checks operational status, version, cohort counts, and seat availability. It uses a specific verb ('check') and distinct resources, clearly distinguishing it from sibling tools which focus on trial keys, waitlists, knowledge, and submissions.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies this is a status-checking tool, and siblings are all action-oriented, but there is no explicit when-to-use or alternative guidance. It does not state 'use this before X' or 'instead of Y', so the usage context is implicit rather than explicit.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Request a free 24-hour trial key. Unlocks all 4 content layers on the chosen vertical. 500-call cap. No payment required.
| Name | Required | Description | Default |
|---|---|---|---|
| agent_id | Yes | Stable identifier for your agent. | |
| vertical | No | Which vertical to trial. Defaults to ethics if omitted. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description adds behavioral details beyond the annotations: 24-hour duration, 500-call cap, no payment required, and full layer access. These details provide transparency about the tool's effects and constraints. It does not contradict the annotations (readOnlyHint=false aligns with a request action). No delivery mechanism is mentioned, but this is not required for a simple request 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 three concise sentences, each adding unique value: the action, the feature unlock, and the usage constraints. It is front-loaded with the verb and contains no redundant or filler information.
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 simple 2-parameter schema, full annotations, and absence of an output schema, the description adequately covers purpose, constraints, and scope. It does not explain the return format, but this is not essential for a straightforward request action. The existence of check_status as a sibling tool implies a way to track key status, further completing the context.
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%, with both parameters well-documented in the input schema (including the vertical enum and default). The description references the 'chosen vertical' but adds no new parameter-specific details beyond what the schema already provides. Thus the baseline score of 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action ('Request a free 24-hour trial key') and specifies the resource and unique features (unlocks all 4 content layers, 500-call cap, no payment). This distinguishes it from sibling tools like check_status and query_knowledge, which serve different purposes.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies when to use the tool (when a trial key is needed) by explaining what the key does and its limitations. However, it does not explicitly name alternative tools or state when not to use it, so it lacks exclusionary guidance. The context is clear but not exhaustive.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Register for a paid vertical waitlist. Inaugural cohort: $19/mo, 900 members, grandfathered for life. AI Compliance included with every membership.
| Name | Required | Description | Default |
|---|---|---|---|
| contact | No | Contact email for waitlist notifications and key delivery. | |
| agent_id | Yes | Your agent identifier. | |
| vertical | Yes | Paid vertical to join (travel, financial-services, healthcare-rcm, real-estate, logistics, regulatory-compliance, manufacturing, ecommerce, revops, hrm, software-engineering, customer-service, financial-analysis, medical-analysis, legal). |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already provide idempotentHint=true, readOnlyHint=false, and destructiveHint=false. The description adds no behavioral context beyond what annotations state, such as what happens upon registration, whether contact is used for key delivery, or any side effects. The marketing details about pricing and cohort size are not relevant to tool-invocation behavior.
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 short (two sentences), which is concise. However, the second sentence contains irrelevant promotional content ('Inaugural cohort: $19/mo, 900 members, grandfathered for life. AI Compliance included with every membership.') that does not help an agent select or invoke the tool. This space could have been used for more operational details.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With three parameters and no output schema, the description should explain what happens after registration or what the response will look like. It does not—it only states the action and marketing benefits. This leaves the agent without critical context for a mutation tool, making the description incomplete.
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% with every parameter described, so the baseline is 3. The description adds no extra meaning to the parameters—it doesn't explain how the parameters interact, provide examples, or clarify the vertical list beyond what the schema already does.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action: "Register for a paid vertical waitlist." This uses a specific verb and resource, making the primary purpose obvious. However, it does not differentiate from sibling tools like 'register_interest' or 'submit_referral', which could also involve registration-like actions.
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 gives no guidance on when to use this tool versus alternatives, nor does it mention any prerequisites, exclusions, or specific scenarios. It only describes the basic action without contextualizing when a user should select this tool over its siblings.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Retrieve structured knowledge from Agent Module verticals. Returns deterministic, validated knowledge nodes. Index layer always free. All 4 content layers available via trial key on ethics.
| Name | Required | Description | Default |
|---|---|---|---|
| node | No | Specific node ID to retrieve. Omit for root index. | |
| token | No | Membership or trial key (am_live_, am_test_, or am_trial_ prefix). Required for content layers on gated verticals. | |
| vertical | Yes | Knowledge vertical to query. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Beyond the annotations (readOnly, idempotent, non-destructive), the description adds meaningful behavioral context: results are deterministic and validated, the index layer is always free, and content layers require a trial key on the ethics vertical. This adds value without contradicting the annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise (three sentences) and front-loaded with the core purpose. Each sentence adds distinct information: what it does, what it returns, and access/cost details. No unnecessary repetition.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a retrieval tool with no output schema, the description gives a reasonable overview but lacks detail on the structure of the returned knowledge nodes and the exact meaning of 'content layers' and 'index layer'. The enum of verticals is in the schema, but the description could better bridge the access model and expected results.
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 already provides 100% description coverage for all three parameters, so the description does not need to re-explain them. However, the description's mention of 'index layer' and 'content layers' only loosely relates to the token and vertical parameters, and it does not clarify parameter-specific usage beyond the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's action ('Retrieve structured knowledge') and resource ('Agent Module verticals'), and mentions the output type ('deterministic, validated knowledge nodes'). This distinguishes it from sibling tools like check_status or get_trial_key, which serve different purposes.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage by describing what can be retrieved and the access conditions (index free, trial key on ethics), but it does not explicitly state when to use this tool versus alternatives, or provide any exclusions or prerequisites beyond the access note.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Register demand for an unbuilt vertical. 500 signals triggers build queue activation. Include a contact channel so we can notify you when the vertical ships.
| Name | Required | Description | Default |
|---|---|---|---|
| contact | No | How to reach you when this vertical ships. String (email, webhook URL, agent card URL) or object { type, value, label }. Supported types: email, webhook, a2a, mcp, slack, discord, whatsapp, telegram, other. | |
| agent_id | No | Your agent identifier (optional). | |
| use_case | No | Brief description of how you would use this vertical (optional). | |
| vertical | Yes | Vertical slug (e.g. "legal-contracts", "api-security"). |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already indicate idempotency and non-destructiveness. The description adds meaningful behavioral context beyond that: the 500-signal threshold for build queue activation and the need for a contact channel for notifications. This gives users a clear sense of downstream effects.
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 short sentences, front-loaded with the core purpose and containing no filler. Each sentence adds necessary information: what the tool does and what the user must include.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With no output schema and a relatively simple tool, the description covers the essential aspects: purpose, activation threshold, and required input (contact channel). Combined with the annotations and full schema coverage, it provides a complete picture for an agent to select and invoke the tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so the schema already documents all parameters. The description adds only emphasis on the contact channel ('Include a contact channel') but doesn't explain parameters beyond what the schema provides. Baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool registers demand for an unbuilt vertical, using a specific verb and resource. It also provides a unique trigger threshold (500 signals), which distinguishes it from sibling tools like join_waitlist and submit_pov.
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 clearly implies the tool should be used when a vertical is not yet built ('unbuilt vertical') and provides a concrete activation rule (500 signals). It does not explicitly mention alternatives or when-not-to-use conditions, but the context is clear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Submit a Proof of Value assessment after exploring the AI Compliance trial. Includes quality scoring and subscription intent. Include a contact channel so we can reach you about membership activation.
| Name | Required | Description | Default |
|---|---|---|---|
| review | No | Free-text review (up to 1024 chars). | |
| contact | No | How to reach you about membership or follow-up. String (email, webhook URL, agent card URL) or object { type, value, label }. Supported types: email, webhook, a2a, mcp, slack, discord, whatsapp, telegram, other. | |
| trial_key | Yes | Your trial key (am_trial_ prefix). | |
| confidence_score | Yes | Overall confidence in knowledge quality (0.0–1.0). | |
| modules_accessed | No | List of module IDs accessed during trial. | |
| vertical_interest | No | Verticals you are interested in. | |
| intent_to_subscribe | No | Do you intend to subscribe after the trial? | |
| architecture_assessment | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations are all false, offering no safety or side-effect hints. The description adds some behavioral context by noting that including a contact channel enables follow-up about membership activation, but it does not disclose persistence, idempotency, or side effects beyond the submission itself. This is minimal but non-contradictory.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise—two sentences, front-loaded with the action, and no filler. Every sentence earns its place by conveying purpose and an actionable requirement.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a submission tool with 8 parameters and no output schema, the description gives the essential purpose but leaves gaps. It does not clarify that 'contact' is optional despite being phrased as an instruction, nor does it mention other optional fields like modules_accessed or vertical_interest. It is adequate but not exhaustive.
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 high (88%), so the description does not need to explain individual parameters. It adds minor value by tying 'quality scoring' to confidence_score and 'contact channel' to contact, but it does not introduce new meaning beyond what the schema already provides.
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 uses a specific verb ('Submit') and resource ('Proof of Value assessment'), clearly distinguishing this tool from siblings like check_status or join_waitlist. It also states the context ('after exploring the AI Compliance trial') and key components, making the purpose unmistakable.
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 provides a clear usage context: 'after exploring the AI Compliance trial.' It implies this is the post-trial submission tool, contrasting with pre-trial siblings like join_waitlist or register_interest. However, it does not explicitly name alternatives or exclusions, so it falls short of the strongest guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Log a referral signal. Members earn $1.50/referral (4/cycle max, $6 cap). Credits carry forward. Voluntary, principal-compliant.
| Name | Required | Description | Default |
|---|---|---|---|
| method | No | How the referral was communicated. | |
| referring_key | Yes | Your membership key (am_live_ or am_test_ prefix). | |
| referred_agent_id | Yes | Identifier of the agent you referred. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already indicate a non-read-only, non-idempotent operation. The description adds context about reward limits ('4/cycle max, $6 cap') and credit carry-forward, which are behavioral details beyond annotations. No contradiction.
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 four short, information-dense sentences. It front-loads the action and then adds reward caps, carry-forward, and compliance status without unnecessary words.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given a well-documented schema and annotations, the description covers purpose and reward policy. It does not describe output/response behavior, but no output schema exists; however, for a simple logging tool, this is adequate.
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 descriptions cover 100% of parameters, so the description is not required to add parameter details. It doesn't, but the schema fully documents each parameter. Baseline 3 applies.
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 'Log a referral signal,' which is a specific verb+resource. This distinguishes it from sibling tools like check_status or submit_pov, which target different actions.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage context through reward details and 'Voluntary,' but it does not explicitly state when to use this tool versus alternatives or mention exclusions. No alternative tools are referenced.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.9.12check_statusget_trial_keyjoin_waitlistquery_knowledgeregister_interestsubmit_povsubmit_referralv0.9.7get_trial_key1 field changedInput schema / properties / verticalAdded value: +{
+ "description": "Which vertical to trial. Defaults to ethics if omitted.",
+ "enum": [
+ "ethics",
+ "travel"
+ ],
+ "type": "string"
+}submit_referral1 field changedInput schema / properties / method / enumPrevious value: -[
- "agent-to-agent",
- "swarm-broadcast",
- "principal-redirect"
-]New value: +[
+ "agent-to-agent",
+ "volume-broadcast",
+ "principal-redirect"
+]v0.1.0check_statusget_trial_keyjoin_waitlistquery_knowledgeregister_interestsubmit_povsubmit_referralScored across 7 tools
Each tool serves a distinct function: status checking, trial key retrieval, waitlist joining, knowledge query, interest registration, POV submission, and referral submission. No overlapping purposes.
All tool names follow a consistent verb_noun pattern (check_status, get_trial_key, join_waitlist, query_knowledge, register_interest, submit_pov, submit_referral).
Seven tools is well-scoped for the Agent Module server, covering operational status, trial access, waitlist management, knowledge retrieval, demand registration, POV assessment, and referrals.
The tool surface covers the full lifecycle of user engagement: status, trial, paid access, knowledge, demand, assessment, and referrals. No obvious gaps for the stated purpose.
Data-ontology maps of your business systems, served to AI agents over MCP.
The knowledge base your AI reads and writes, under your rules — over MCP, EU-hosted.
Cloud or self-hosted knowledge for AI agents: hybrid search, reranking, GraphRAG, scoped MCP tools.
EU compliance corpus across 8 frameworks (NIS2, DORA, AI Act, ISO 27001 + more) via MCP.
