Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@@promptjang/mcp-server Send a test event to target 'demo' with payload 'ping'"
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.
Official local MCP server for PromptJang. It exposes bounded delivery and mailbox tools over stdio.
Targets and delivery: list_targets, send_event, get_event, replay_event, get_usage
Agent mailbox: list_mailboxes, list_unread, get_message, claim_message, ack_message
The server never accepts a destination URL. send_event requires a registered target ID returned by list_targets.
MCP-originated items follow the same usage contract as the application API. Each newly accepted webhook event, mailbox message, or A2A handoff creates one usage record at the published $0.10 per 1,000 accepted-item rate. Retries, reads, claims, acknowledgements, replays, rejected requests, and idempotent duplicates create no additional usage. Live billing is disabled during the founder-assisted beta, so these records are not currently exported to Stripe or charged.
Mailbox consumers follow one bounded lifecycle: list unread messages, claim one for a five-minute lease, process it, then acknowledge it with the same API key.
Requires Node.js 20 or later and a PromptJang API key.
export PROMPTJANG_API_KEY=pj_live_your_key
npx --yes github:PromptJang/promptjang-mcpThe source repository is the supported install path for v0.1.1. The
@promptjang/mcp-server npm registry package is not published yet.
Related MCP server: PromptDB MCP Server
[mcp_servers.promptjang]
command = "npx"
args = ["-y", "github:PromptJang/promptjang-mcp"]
[mcp_servers.promptjang.env]
PROMPTJANG_API_KEY = "pj_live_your_key"claude mcp add promptjang --env PROMPTJANG_API_KEY=pj_live_your_key -- npx -y github:PromptJang/promptjang-mcpKeep the API key in an environment or secret manager. Do not commit it to a project configuration.
Acknowledge a claimed mailbox message after successful processing. Only the API key that owns the active claim can acknowledge it.
| Name | Required | Description | Default |
|---|---|---|---|
| message_id | Yes | PromptJang 32-character event or target identifier | |
| response_format | No | Human-readable markdown or machine-readable JSON | markdown |
| Name | Required | Description |
|---|---|---|
| id | Yes | |
| status | Yes | |
| acknowledged_at | No | |
| claim_expires_at | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already signal mutation (readOnlyHint=false), idempotency, and non-destructiveness. The description adds value by disclosing the active-claim ownership requirement and the expected timing. No contradiction, though the exact effect of acknowledgement on mailbox state is not spelled out.
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 concise sentences with no filler: the first gives the action and timing, the second adds the single gating condition. It does not repeat schema fields or annotation data.
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 simple two-parameter tool with an output schema and full annotations, the description covers when to call and who is allowed to call. The output schema handles return-value details, so nothing essential is missing for correct invocation.
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 description coverage is 100%, with message_id format/length and response_format enum/default already documented. The description adds no parameter-specific meaning, so the schema appropriately carries the burden.
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 names a concrete action ('acknowledge') and a specific resource ('claimed mailbox message'), with 'after successful processing' establishing intent. 'Acknowledge' clearly contrasts with sibling claim_message and read-oriented tools like get_message and list_unread, so an agent can distinguish it.
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 states the lifecycle condition—use after successful processing—and the ownership prerequisite: only the API key that owns the active claim can acknowledge. It does not enumerate all sibling alternatives, but the claim-then-acknowledge sequencing is unambiguous.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Atomically claim one unread mailbox message for five minutes. The same PromptJang API key must acknowledge it after processing.
| Name | Required | Description | Default |
|---|---|---|---|
| message_id | Yes | PromptJang 32-character event or target identifier | |
| response_format | No | Human-readable markdown or machine-readable JSON | markdown |
| Name | Required | Description |
|---|---|---|
| id | Yes | |
| status | Yes | |
| acknowledged_at | No | |
| claim_expires_at | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The annotations only say the operation is not read-only, is idempotent, and not destructive. The description adds substantive behavioral facts: atomicity, a five-minute claim duration, and the same-key acknowledgement requirement. It stops short of detailing expiration/recovery or failure behavior, but goes well beyond 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?
Two sentences with no filler: the first states the action and lease, the second states the required follow-up. The most important constraint (acknowledge with the same key) is placed at the end but earned through context.
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 two-parameter tool with a full schema and output schema, the description provides the behavioral essentials: what is claimed, for how long, and what obligation the caller incurs. Nothing critical is missing for an agent to invoke it correctly in the claim/process/ackworkflow.
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 description coverage is 100%, and both message_id and response_format already have descriptions and defaults/enum values. The tool description adds no parameter-level detail beyond mentioning the claimed message, so the 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 names a specific action ('claim'), a specific resource ('one unread mailbox message'), and adds a concrete constraint ('for five minutes'), so an agent knows exactly what operation is offered. This is clearly distinct from siblings like list_unread, get_message, and ack_message.
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?
It conveys the workflow context: a message should be claimed before processing and then acknowledged with the same API key, which implies when this tool fits relative to siblings. It does not explicitly enumerate alternative tools or exclusions, such as 'for read-only access use get_message', but the claim/ack sequence is sufficiently clear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Read one PromptJang event and its delivery attempts by event ID. Use this to inspect current state, response codes, timing, retry count, and the last delivery error.
| Name | Required | Description | Default |
|---|---|---|---|
| event_id | Yes | PromptJang 32-character event or target identifier | |
| response_format | No | Human-readable markdown or machine-readable JSON | markdown |
| Name | Required | Description |
|---|---|---|
| event | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true and idempotentHint=true, so the safety profile is well covered. The description adds useful behavioral context beyond annotations by specifying that the tool returns delivery attempts and last delivery error, which clarifies the operational depth of the read operation.
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 with no filler. It front-loads the core operation and selection key, then immediately states the inspection use case. Every clause contributes useful 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 simple single-record read tool, the description is complete. It explains what is returned, how to select it, and why an agent would use it. The rich annotations and existing output schema cover the remaining concerns like idempotency and return structure.
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 description coverage is 100%, so both parameters (event_id and response_format) are already fully documented by the schema. The description reinforces that the lookup is 'by event ID' but adds no new semantic details about the parameters themselves, matching the baseline for full 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 states a specific verb ('Read'), a clear resource ('one PromptJang event and its delivery attempts'), and a precise selector ('by event ID'). It also lists the particular data points available (state, response codes, timing, retry count, last delivery error), which helps distinguish it from sibling tools like list_unread or get_message.
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 clear context for when to use the tool: 'Use this to inspect current state, response codes, timing, retry count, and the last delivery error.' It does not explicitly name alternatives or exclusions, but the use case is specific enough that an agent can determine when this tool is appropriate.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Read one mailbox message and its JSON payload without claiming or acknowledging it.
| Name | Required | Description | Default |
|---|---|---|---|
| message_id | Yes | PromptJang 32-character event or target identifier | |
| response_format | No | Human-readable markdown or machine-readable JSON | markdown |
| Name | Required | Description |
|---|---|---|
| message | Yes | |
| payload | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already provide a rich safety profile (readOnlyHint, idempotentHint, non-destructive), so the bar is lower. The description adds genuine value beyond those hints by disclosing that reading does not trigger the mailbox's claim/acknowledgment state machine — a subtle workflow behavior a generic readOnlyHint would not convey. No contradiction with 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?
A single 13-word sentence front-loads the verb and resource, then spends its remaining words on the critical differentiator ('without claiming or acknowledging it'). There is zero filler, and every word earns its place.
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 simple single-message read tool, the definition is complete: annotations cover the safety profile, the output schema covers return shape, the schema covers both parameters, and the description covers the one non-obvious behavioral nuance. Nothing an agent needs to invoke this correctly is missing.
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 description coverage is 100%, so both message_id and response_format are already fully documented with type, pattern, enum, and description. The description's mention of 'JSON payload' slightly reinforces the response_format parameter but adds no meaning beyond what the schema provides, so the 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 uses a specific verb ('Read') with a precise resource ('one mailbox message and its JSON payload') and immediately distinguishes itself from claim_message and ack_message by stating it operates 'without claiming or acknowledging.' This makes the tool's purpose unambiguous and easily separable from all nine siblings without opening the schema.
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 phrase 'without claiming or acknowledging it' gives clear context that this is the side-effect-free inspection tool, implicitly routing claim_message and ack_message use cases elsewhere. It does not explicitly name those alternatives or state when-not-to-use conditions, but the context is strong enough that an agent can select correctly.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Read the current calendar-month accepted-item count, estimated USD charge, and pending meter-event count. Retries and replays are excluded. During the billing-disabled founder beta, usage remains local and exempt workspaces report a zero charge estimate.
| Name | Required | Description | Default |
|---|---|---|---|
| response_format | No | Human-readable markdown or machine-readable JSON | markdown |
| Name | Required | Description |
|---|---|---|
| period_end | Yes | |
| period_start | Yes | |
| accepted_events | Yes | |
| estimated_cents | Yes | |
| estimated_units | Yes | |
| pending_meter_events | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare read-only, idempotent, and non-destructive behavior. The description adds valuable behavioral context beyond annotations: retries and replays are excluded from counts, and the billing-disabled founder beta behavior is disclosed (local usage, zero charge estimate for exempt workspaces).
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 dense sentences front-load the main purpose before adding important exclusions and beta caveats. Every sentence adds useful information and there is no filler or repetition of schema 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?
For a simple read-only usage tool with an output schema, the description covers the core return values, exclusions, and special-case behavior. No additional prerequisites, auth notes, or alternative routing are needed given the tool's scope and sibling set.
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 only parameter, response_format, is fully documented in the schema with an enum and default value. Schema description coverage is 100%, so the description does not need to repeat it. The baseline of 3 applies since the schema carries the semantic load.
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 the specific verb 'Read' and names the exact metrics: accepted-item count, estimated USD charge, and pending meter-event count. It clearly identifies the resource as current calendar-month usage, which distinguishes it from the sibling message/event/mailbox tools.
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 clear context: it is for reading usage/billing status for the current month. It does not explicitly name alternatives or exclusions, but the sibling tools are all unrelated to usage/billing, so there is no real ambiguity about when to use this tool.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List durable agent mailboxes available to this API key. Returns registered mailbox IDs for list_unread and send_event.
| Name | Required | Description | Default |
|---|---|---|---|
| response_format | No | Human-readable markdown or machine-readable JSON | markdown |
| Name | Required | Description |
|---|---|---|
| count | Yes | |
| mailboxes | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already establish readOnly, idempotent, and non-destructive behavior. The description adds useful context beyond that: mailboxes are 'durable', scoped to the API key, and the returned IDs serve list_unread and send_event. This supports correct usage without contradicting 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?
Two concise sentences, both earning their place: the first defines the resource and scope, the second clarifies the return value and downstream consumers. No filler or 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 simple list tool with one optional fully-documented parameter, an output schema, and comprehensive annotations, the description is complete. An agent can call it reliably without additional assumptions or missing 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 description coverage is 100%, and the single optional parameter response_format has an enum, default, and description. The tool description does not add parameter-specific meaning, but the schema already carries the full burden, so 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 uses a specific verb ('List') with a clear resource ('durable agent mailboxes available to this API key') and states that it returns registered mailbox IDs for list_unread and send_event. This clearly distinguishes it from sibling tools like list_targets and list_unread, which operate on different resources or 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 when to use the tool: to discover mailbox IDs needed by list_unread and send_event. It provides clear usage context, though it does not explicitly state when not to use it or mention alternatives, so it stops just short of full guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List registered PromptJang webhook, mailbox, and A2A targets available to this API key. Call this before send_event; this tool never creates or changes targets.
| Name | Required | Description | Default |
|---|---|---|---|
| response_format | No | Human-readable markdown or machine-readable JSON | markdown |
| Name | Required | Description |
|---|---|---|
| count | Yes | |
| targets | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, idempotentHint, and destructiveHint false; the description reinforces this with 'never creates or changes targets' and adds that results are scoped to the API key. This goes beyond the structured annotations without contradicting them.
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 with no wasted words. The core action is front-loaded, and the usage note is placed immediately after, making the definition easy to parse quickly.
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?
This is a simple, safe listing tool with one optional enum parameter, a full output schema, and strong annotations. The description adds workflow context and clarifies scope, so nothing necessary is missing.
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 only parameter, response_format, is fully described in the schema with its enum values and default. The description adds no parameter-specific detail, but with 100% schema description coverage this is acceptable; 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?
Description states a specific verb ('List') and resource ('PromptJang webhook, mailbox, and A2A targets'), and scopes it to the API key. Enumerating target types makes the tool's broad scope clear and implicitly distinguishes it from the narrower sibling list_mailboxes.
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?
Explicitly instructs the agent to 'Call this before send_event,' which gives clear workflow context. It does not explicitly name when not to use this tool or mention list_mailboxes as an alternative, but the primary usage signal is strong.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List unread messages in one registered PromptJang mailbox without claiming or changing them.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | ||
| mailbox_id | Yes | Mailbox target ID returned by list_mailboxes | |
| response_format | No | Human-readable markdown or machine-readable JSON | markdown |
| Name | Required | Description |
|---|---|---|
| count | Yes | |
| messages | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, idempotentHint, and destructiveHint=false. The description adds value by clarifying that messages are not claimed or changed, which is operationally meaningful beyond generic read-only, and by narrowing scope to unread messages in a single mailbox. It does not contradict 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?
A single, tightly scoped sentence with no filler. It front-loads the action and resource, then adds the crucial no-side-effect qualifier, making every word earn its place.
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 output schema, annotations, and schema-described parameters, the description is nearly complete for a simple read-only listing tool. The only real gap is explicit alternative-tool routing, which is a usage-guidelines deficit rather than a completeness failure for invoking the tool correctly.
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 description coverage is 67%: mailbox_id and response_format are already documented in the schema, and limit has clear type/default/min/max constraints. The description adds no per-parameter detail beyond 'one registered mailbox', reinforcing mailbox_id but not explaining limit or response format further. This is acceptable but not additive.
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 states a specific verb ('List'), a clear resource ('unread messages in one registered PromptJang mailbox'), and an explicit non-side-effect scope ('without claiming or changing them'). This distinguishes it from siblings like claim_message and ack_message, and from get_message, which targets a single message.
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 for read-only enumeration of unread messages and explicitly says it does not claim or change messages, which helps rule out claim/ack tools. However, it does not explicitly state when to use this versus list_mailboxes or get_message, or name alternative tools for those cases.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Create a non-billable replay of an existing PromptJang event using its stored payload and registered target. Returns the new replay event ID. This changes delivery state but does not delete or overwrite the source event.
| Name | Required | Description | Default |
|---|---|---|---|
| event_id | Yes | PromptJang 32-character event or target identifier | |
| response_format | No | Human-readable markdown or machine-readable JSON | markdown |
| Name | Required | Description |
|---|---|---|
| id | Yes | |
| status | Yes | |
| idempotent_replay | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already indicate this is a non-read-only, non-idempotent, open-world action. The description adds valuable context by stating that it is non-billable, changes delivery state, and does not delete or overwrite the source event. This meaningfully supplements the annotations without contradicting them.
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 with no redundant phrasing. It front-loads the core action and result, then adds the key side-effect clarification. Every sentence earns its place.
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 an output schema present, the description does not need to detail return values beyond noting the new replay event ID. It covers the key behavioral context: non-billable, non-destructive, delivery-state-changing, and source-preserving. Minor gaps like prerequisites or failure cases are not significant enough to lower further.
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 description coverage is 100%, so the input schema already documents event_id and response_format. The tool description does not add parameter-level detail beyond referencing an 'existing PromptJang event', which maps naturally to event_id. Baseline 3 is appropriate given full 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 states a specific action ('Create a non-billable replay') on a specific resource ('existing PromptJang event') with a clear method ('using its stored payload and registered target'). It also differentiates from siblings like send_event by emphasizing 'non-billable' and 'existing event'.
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 when to use the tool: when an existing PromptJang event needs to be re-delivered without billing, using its stored payload. It does not explicitly name alternatives or state when not to use it, so it stops short of a 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Accept one JSON item for a registered webhook, mailbox, or A2A target. Requires target_id from list_targets and never accepts an arbitrary URL. Set idempotency_key when repeating the same logical item.
| Name | Required | Description | Default |
|---|---|---|---|
| payload | Yes | JSON object to deliver; maximum serialized size is 256 KB | |
| target_id | Yes | Registered target ID returned by list_targets | |
| event_type | No | Optional event type such as payment.succeeded | |
| idempotency_key | No | Stable key that prevents duplicate delivery and billing | |
| response_format | No | Human-readable markdown or machine-readable JSON | markdown |
| Name | Required | Description |
|---|---|---|
| id | Yes | |
| status | Yes | |
| idempotent_replay | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations mark the operation as non-readonly and non-idempotent, while the description adds meaningful boundaries: only registered targets are accepted, arbitrary URLs are never allowed, and idempotency_key should be set when re-sending the same logical item. Delivery/error behavior isn't detailed, but the annotation burden is already partly met.
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?
Three concise, front-loaded sentences: purpose, key constraint, and idempotency guidance. Every sentence carries useful signal with no fluff.
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 five-param send operation with an output schema and annotations, the description plus schema is sufficient for an agent to call correctly. Minor gaps about delivery failure behavior and explicit alternative routing remain but don't block correct use.
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 has 100% description coverage, including target_id provenance and idempotency_key's duplicate-prevention role. The description mostly restates these schema details rather than adding new parameter meaning.
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?
States a specific verb and resource: accepts one JSON item for a registered webhook, mailbox, or A2A target. It clearly distinguishes from sibling read/replay tools by focusing on sending a new event and requiring a registered target_id from list_targets.
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?
Provides concrete invocation context: target must be pre-registered, target_id must come from list_targets, and arbitrary URLs are rejected. It doesn't explicitly name alternatives or when-not-to-use cases, but the context is clear enough for correct routing.
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.
v0.1.1ack_messageclaim_messageget_eventget_messageget_usagelist_mailboxeslist_targetslist_unreadreplay_eventsend_eventScored across 10 tools
Each tool targets a distinct resource or action, and the descriptions make the boundaries clear. The only mild overlap is between list_targets and list_mailboxes, since mailbox targets and durable mailboxes are related, but the descriptions differentiate them well.
Tool names consistently follow a verb_noun snake_case pattern: list_*, get_*, claim_*, ack_*, send_*, replay_*. This makes the set highly predictable and easy to navigate.
Ten tools is well-scoped for an event, mailbox, and usage monitoring server. Each tool serves a clear purpose without redundancy or bloat.
The core lifecycle is covered: send events, inspect delivery, replay failed events, list/read/claim/acknowledge mailbox messages, and check usage. A minorgap is the lack of a list-events endpoint to browse past events without knowing their IDs, but agents can work around this.
Read a project's prompts, logs and agents, and send new work to the agent on your own machines.
Manage Speko voice-AI agents, sessions, calls, phone numbers, knowledge bases, evals, and docs.
Manage portable AI agent playbooks, Agent Skills, MCP configurations, personas, and memory.
Illumora Craft — evidence-augmented prompt compile for agents via remote MCP (OAuth) or local stdio.
