Sequence sentinel: wall events + channel-switch scoring; HALT, hold, signed receipt.
Scored across 6 tools
The hold lifecycle tools (list_holds, deny_hold, release_hold) are clearly distinct. However, score_action, query_verdict, and observe_result all touch scoring/verdict/result recording, and the cryptic rail/x402 descriptions make their boundaries less obvious. An agent could plausibly misselect score_action when it actually needs query_verdict.
All tool names follow a consistent snake_case verb_noun pattern: deny_hold, list_holds, observe_result, query_verdict, release_hold, score_action. The only minor variation is pluralization in list_holds, which is natural and does not break the pattern.
Six tools is well-scoped for a focused sentinel that manages holds, scores actions, observes runtime results, and queries verdicts. Each tool appears to earn its place without obvious padding or severe under-coverage.
The set covers hold listing, denial, and release, plus action scoring, result observation, and verdict querying. Minor gaps remain around creating holds, inspecting wall configuration/state, or retrieving a single verdict/hold by identifier, but core workflows appear workable.
Deny a held action.
| Name | Required | Description | Default |
|---|---|---|---|
| hold_id | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden of disclosing behavior. "Deny" implies a state-changing action, but the description says nothing about permanence, side effects, permissions, or what happens to the held action.
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 front-loaded sentence with no filler or repetition. It is efficient, though its brevity contributes to the broader specification gaps.
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 annotations, no output schema, and no parameter documentation, the definition is far too thin for a state-changing operation. It does not explain consequences, expected result, or failure conditions.
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 single required parameter, hold_id, has 0% schema description coverage, and the description adds no format or source information. "Held action" only loosely implies that hold_id identifies the hold, which is minimal compensation.
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 clear verb ("Deny") and object ("a held action"), so the broad operation is inferable. However, "held action" is not defined, and it does not distinguish this tool from siblings such as release_hold or score_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?
There is no guidance on when to deny versus release a hold, what preconditions apply, or what happens after denial. The agent is left to infer usage entirely from the one-line name.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List your held actions with wall + signals.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full behavioral burden: it does not state that this is a read-only operation, whether results are paginated or filtered, whether auth/scope is required, or what the entries contain. "wall + signals" hints at returned fields but is too cryptic to count as disclosure.
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 short sentence with no filler, and the core action is front-loaded. The unexplained "wall + signals" phrase spends words without earning them, which keeps it out of the top band.
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 zero-param list tool with no output schema and no annotations, so the description is the only source of information about what comes back — yet "wall + signals" is never defined. An agent cannot know what a "hold", a "wall", or a "signal" looks like in the response.
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 tool takes zero parameters and the schema is an empty object at 100% coverage, so there is nothing for the description to document. Baseline 4 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?
"List your held actions" gives a specific verb (list) and resource (held actions), which distinguishes it from deny_hold/release_hold siblings. However, "with wall + signals" is unexplained internal jargon, so the scope of what is listed is only partially clear.
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?
There is no statement of when to use this versus the related siblings (release_hold, deny_hold, observe_result), nor any prerequisites or exclusions. The agent must infer the usage context entirely.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Record a runtime tool result; a wall is recorded only when the event is runtime-sourced.
| Name | Required | Description | Default |
|---|---|---|---|
| event | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full behavioral burden. It discloses only that a 'wall' is recorded when the event is runtime-sourced, but never explains what a wall is, what happens otherwise, or whether recording has side effects, persistence, or error conditions.
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 front-loaded sentence with no wasted clauses. Its terseness reflects under-specification rather than poor conciseness.
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 tool with a required nested object, no schema descriptions, no annotations, and no output schema, the description leaves major gaps. It does not explain the event shape, the meaning of 'wall,' or how this operation relates to sibling tools.
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 0% for the single required 'event' parameter, which is an undocumented nested object. The description adds that the event is a runtime tool result and that runtime-sourced status determines wall recording, but it does not describe the event's expected fields or structure.
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 first clause states a clear verb and resource: 'Record a runtime tool result.' It does not differentiate from siblings such as score_action or query_verdict, and the undefined term 'wall' adds ambiguity, but the core purpose is identifiable.
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 the hold, verdict, or scoring siblings. The runtime-sourced condition is a behavioral rule, not usage guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Per-verdict query shape over MCP: scores via the tier rail; the $0.01 x402 rail is POST /query.
| Name | Required | Description | Default |
|---|---|---|---|
| action | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description must carry the full behavioral burden, but it omits permissions, side effects, return shape, and whether the operation is read-only. The mention of a $0.01 x402 rail and POST /query adds some cost/transport context but not enough to characterize the tool's 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?
It is a single sentence and not bloated, but the jargon makes it hard to parse and the core purpose is not front-loaded. The x402 endpoint detail may not help an MCP agent select the tool.
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 opaque nested 'action' parameter, no annotations, and no output schema, the description is not complete enough for correct invocation. It gives some rail/cost context but leaves the action contract and return behavior unclear.
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 schema has one required 'action' object with 0% description coverage, and the description never explains what 'action' contains or what valid values look like. With low schema coverage, the description should compensate, but it provides no parameter semantics.
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 internal jargon such as 'per-verdict query shape,' 'tier rail,' and 'x402 rail' instead of a clear verb+resource statement. It hints that the tool queries verdicts or scores, but does not plainly state what it returns or how it differs from siblings.
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 when-to-use guidance, prerequisites, or comparison to sibling tools like score_action or observe_result. The only routing information concerns transport rails, not when an agent should choose this MCP tool.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Release a hold (single-use; returns exactly the held action).
| Name | Required | Description | Default |
|---|---|---|---|
| hold_id | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full behavioral burden. It usefully discloses that the operation is single-use and returns the held action, but it omits permissions, side effects, irreversibility, and error behavior for invalid or already-released holds.
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 front-loaded sentence with no wasted words, and the key action and return behavior are stated immediately. The parenthetical phrase "held action" is compact but slightly ambiguous, which keeps it from a 5.
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 one-parameter tool, the description covers the core action and a return characteristic, and no output schema exists to explain the return value further. However, with no annotations and low schema coverage, it remains incomplete regarding authorization, side effects, and failure modes.
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 schema has only hold_id with 0% description coverage, so the description should compensate by explaining what the identifier represents or how it is obtained. It does not add meaning beyond the parameter name and the general instruction to release a hold.
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 clear verb and resource ("Release a hold") and adds useful scope with "single-use; returns exactly the held action." It does not explicitly distinguish itself from siblings such as deny_hold or list_holds, so it falls short of a 5.
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?
There is no explicit guidance on when to use this tool versus deny_hold, list_holds, or the other siblings. The description implies releasing a known hold but does not state preconditions, alternatives, or when-not-to-use conditions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Score one action against the operator's live wall state (tier allowance accounting applies).
| Name | Required | Description | Default |
|---|---|---|---|
| action | Yes |
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 hints that tier allowance accounting is involved, which suggests a resource-consuming or state-affecting operation, but it never says whether state is mutated, whether the call is idempotent, what happens when allowance is exhausted, or what authorization is required.
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 sentence with no filler, and the core operation is front-loaded before the parenthetical qualifier. The jargon costs some clarity but there is no padding.
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, no annotations, and an opaque nested parameter, the description would need to explain the return value (a score? a verdict?) and the state effects. It supplies only a fragment of that, so an agent cannot reliably call this 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?
The single required parameter 'action' is an unconstrained object with no properties and 0% schema description coverage, and the description adds nothing about its expected shape or fields. For a nested, undocumented object, the description does not compensate for the schema gap at all.
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 verb+resource ('score one action') is concrete and the scope ('against the operator's live wall state') narrows it, so it is more than a restatement of the name. However, the domain terms 'live wall state' and 'tier allowance accounting' are undefined, leaving the actual effect of 'scoring' ambiguous. It also does not distinguish itself from siblings like query_verdict or observe_result.
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?
There is no statement of when to use this tool versus query_verdict, observe_result, or the hold tools. No prerequisites or preconditions are given, only the parenthetical that accounting applies. The agent must infer usage entirely from the name.
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.
deny_holdlist_holdsobserve_resultquery_verdictrelease_holdscore_actionSterile phase for irreversible agent actions: frozen intent, deferred inputs, signed exit receipt.
Evidence-bound stopping for agents: judgment ledger, binding futility verdicts, signed receipts.
Real MCP handshake witness graded per-client, signed expiring receipt.
Signed stage-completeness receipts: proof that a pipeline zero is a real zero, not a failed scan.
Add one secure layer between your agents and this server.