Kevros — Agentic Identity Trust
Agentic identity trust: precision decisioning, cryptographic release tokens, hash-chained proof
我该使用它吗
质量与安全性
A
发现（3）
- LOWTool 'verify-outcome' doesn't follow camelCase/snake_case在 verify-outcome 中
- LOWTool 'check-peer' doesn't follow camelCase/snake_case在 check-peer 中
- LOWTool 'verify-token' doesn't follow camelCase/snake_case在 verify-token 中
基于对工具定义和协议合规性的自动分析。
上下文开销
~2,136token 数（工具定义）
~2.1 KB典型响应大小
对注意力有中等影响（占 128k 上下文窗口的 1.67%）
这是每次将服务器的工具加载到模型上下文窗口时所消耗的大致 token 数。数值越高，可用于其他任务的注意力就越少。
安装
一键安装
将以下内容添加到你的 `claude_desktop_config.json` 文件中：
{
"mcpServers": {
"kevros": {
"url": "https://governance.taskhawktech.com/mcp/"
}
}
}远程端点
https://governance.taskhawktech.com/mcp/streamable-http它能做什么
工具清单
工具（9）
🟢 只读🟡 写入🔴 删除⚪ 未知
⚪verify(action_type, action_payload, agent_id, policy_context, template_id, ...)
Verify an action against policy bounds before executing it. Returns ALLOW (proceed), CONSTRAIN (proceed with modified values; emitted on the wire as the legacy value CLAMP - both refer to the same verdict, CONSTRAIN is the public name), or DENY (stop). Every verification is recorded in a hash-chained provenance ledger. Cost: $0.01 per call.
输入模式
{
"type": "object",
"properties": {
"action_type": {
"description": "Category of action to verify (e.g. 'deploy', 'trade', 'send_email')",
"type": "string"
},
"action_payload": {
"additionalProperties": true,
"description": "Action parameters to verify against policy bounds",
"type": "object"
},
"agent_id": {
"description": "Unique identifier of the agent requesting verification",
"type": "string"
},
"policy_context": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Optional policy overrides such as max_values or forbidden_keys"
},
"template_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "Named policy template (e.g. 'robotics-arm', 'financial-transaction'). Merged with policy_context."
},
"idempotency_key": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "Optional key for idempotent retries - same key returns cached result"
}
},
"required": [
"action_type",
"action_payload",
"agent_id"
],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"result": {
"type": "string"
}
},
"required": [
"result"
],
"x-fastmcp-wrap-result": true
}🟡attest(agent_id, action_description, action_payload, context)
Create a hash-chained provenance record for an action you've taken. Each attestation extends the append-only evidence chain. The hash can be independently verified by any third party. Cost: $0.02 per call.
输入模式
{
"type": "object",
"properties": {
"agent_id": {
"description": "Unique identifier of the attesting agent",
"type": "string"
},
"action_description": {
"description": "Human-readable description of the action taken",
"type": "string"
},
"action_payload": {
"additionalProperties": true,
"description": "Structured data describing the action (hashed into provenance)",
"type": "object"
},
"context": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Optional metadata such as environment, trigger, or session info"
}
},
"required": [
"agent_id",
"action_description",
"action_payload"
],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"result": {
"type": "string"
}
},
"required": [
"result"
],
"x-fastmcp-wrap-result": true
}⚪bind(agent_id, intent_type, intent_description, command_payload, goal_state, ...)
Declare an intent and cryptographically bind it to a command. Proves that the command was issued in service of the declared intent. Use verify-outcome after execution to close the loop. Cost: $0.02 per call.
输入模式
{
"type": "object",
"properties": {
"agent_id": {
"description": "Unique identifier of the agent declaring intent",
"type": "string"
},
"intent_type": {
"description": "Category of intent (e.g. 'navigation', 'transaction', 'deployment')",
"type": "string"
},
"intent_description": {
"description": "Human-readable description of what the agent intends to do",
"type": "string"
},
"command_payload": {
"additionalProperties": true,
"description": "The command that will be executed to fulfill this intent",
"type": "object"
},
"goal_state": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Expected end state for outcome verification"
},
"intent_source": {
"default": "AI_PLANNER",
"description": "Origin of intent: AI_PLANNER, HUMAN_OPERATOR, or SYSTEM",
"type": "string"
},
"parent_intent_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "ID of parent intent for hierarchical intent chains"
}
},
"required": [
"agent_id",
"intent_type",
"intent_description",
"command_payload"
],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"result": {
"type": "string"
}
},
"required": [
"result"
],
"x-fastmcp-wrap-result": true
}⚪verify-outcome(agent_id, intent_id, binding_id, actual_state, tolerance)
Verify that an executed action achieved its declared intent. Closes the loop: intent -> command -> action -> outcome -> verification. Free (included with bind).
输入模式
{
"type": "object",
"properties": {
"agent_id": {
"description": "Unique identifier of the agent whose outcome is being verified",
"type": "string"
},
"intent_id": {
"description": "ID of the original intent from governance_bind",
"type": "string"
},
"binding_id": {
"description": "ID of the intent-command binding from governance_bind",
"type": "string"
},
"actual_state": {
"additionalProperties": true,
"description": "Observed end state after action execution, compared against goal_state",
"type": "object"
},
"tolerance": {
"default": 0.1,
"description": "Numeric tolerance for goal matching (0.1 = 10% deviation allowed)",
"type": "number"
}
},
"required": [
"agent_id",
"intent_id",
"binding_id",
"actual_state"
],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"result": {
"type": "string"
}
},
"required": [
"result"
],
"x-fastmcp-wrap-result": true
}🟢bundle(agent_id, time_range_start, time_range_end, max_records, include_intent_chains, ...)
Generate a certifier-grade compliance evidence bundle. Contains hash-chained provenance, intent bindings, PQC attestations, and verification instructions. Independently verifiable without Kevros access. Cost: $0.05 per call.
输入模式
{
"type": "object",
"properties": {
"agent_id": {
"description": "Agent whose provenance records to include in the bundle",
"type": "string"
},
"time_range_start": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "ISO 8601 start time filter (inclusive)"
},
"time_range_end": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "ISO 8601 end time filter (inclusive)"
},
"max_records": {
"default": 10000,
"description": "Maximum number of provenance records to include",
"type": "integer"
},
"include_intent_chains": {
"default": true,
"description": "Include intent-command binding chains in the bundle",
"type": "boolean"
},
"include_pqc_signatures": {
"default": true,
"description": "Include post-quantum ML-DSA-87 block signatures",
"type": "boolean"
},
"include_verification_instructions": {
"default": true,
"description": "Include step-by-step verification procedure for auditors",
"type": "boolean"
}
},
"required": [
"agent_id"
],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"result": {
"type": "string"
}
},
"required": [
"result"
],
"x-fastmcp-wrap-result": true
}🟢health(verbose)
Check the governance gateway health status. Free.
输入模式
{
"type": "object",
"properties": {
"verbose": {
"default": false,
"description": "Return additional details such as chain length and PQC signing status",
"type": "boolean"
}
},
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"result": {
"type": "string"
}
},
"required": [
"result"
],
"x-fastmcp-wrap-result": true
}🟢status(include_chain_details)
Check your current usage and quota: calls used, calls remaining, tier, rate limits, and billing status. Free.
输入模式
{
"type": "object",
"properties": {
"include_chain_details": {
"default": false,
"description": "Include hash-chain integrity check and latest provenance epoch",
"type": "boolean"
}
},
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"result": {
"type": "string"
}
},
"required": [
"result"
],
"x-fastmcp-wrap-result": true
}🟢check-peer(agent_id)
Check another agent's trust score and governance history. Returns trust score (0-100), chain length, attestation count, and tier. Free, no API key needed.
输入模式
{
"type": "object",
"properties": {
"agent_id": {
"description": "ID of the peer agent to look up",
"type": "string"
}
},
"required": [
"agent_id"
],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"result": {
"type": "string"
}
},
"required": [
"result"
],
"x-fastmcp-wrap-result": true
}🟢verify-token(release_token, token_preimage)
Verify a release token from another agent. Confirms the token is authentic, was issued by the Kevros gateway, and remains currently authorized after any halt or mode transition. Free, no API key needed.
输入模式
{
"type": "object",
"properties": {
"release_token": {
"description": "Release token string received from governance_verify",
"type": "string"
},
"token_preimage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "Deprecated. The gateway now performs verification entirely server-side using the release_token alone; any value supplied here is ignored. Field retained for backward compatibility with v1 SDK callers and will be removed in a future release."
}
},
"required": [
"release_token"
],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"result": {
"type": "string"
}
},
"required": [
"result"
],
"x-fastmcp-wrap-result": true
}社区
证据
最近观测
已验证未记录版本9 个工具
已验证未记录版本9 个工具
已验证未记录版本9 个工具