portkey-admin-mcp

get_cache_hit_latency

Read-onlyIdempotent

Retrieve cache-hit-only latency time-series data with total and average latency per bucket. Measure cached-response speed and pinpoint performance bottlenecks using filters for time range, configs, models, and more.

Instructions

Get cache-hit-only latency time-series data with summary.total_latency, summary.avg_latency, and per-bucket total/avg latency. Use this to evaluate cached-response speed; use get_latency_analytics for all requests. Enterprise-gated. Returns 403 on non-Enterprise Portkey plans.

Input Schema

TableJSON Schema

Name	Required	Description
`configs`	No	Legacy Portkey query param for config slugs. Comma-separated string; prefer config_slugs for structured inputs.
`span_id`	No	Legacy Portkey query param for span IDs. Comma-separated string; prefer span_ids for structured inputs.
`cost_max`	No	Maximum cost in cents to filter by
`cost_min`	No	Minimum cost in cents to filter by
`metadata`	No	Legacy Portkey query param for metadata filtering. Stringified JSON object, e.g. '{"env":"prod","app":"myapp"}'; prefer metadata_filter for structured inputs.
`span_ids`	No	Structured alias for span_id. Use an array of span IDs; normalized to the legacy comma-separated Portkey query param.
`trace_id`	No	Legacy Portkey query param for trace IDs. Comma-separated string; prefer trace_ids for structured inputs.
`trace_ids`	No	Structured alias for trace_id. Use an array of trace IDs; normalized to the legacy comma-separated Portkey query param.
`api_key_ids`	No	Legacy Portkey query param for API key UUIDs. Comma-separated string; request_analytics also accepts an array and normalizes it to this form.
`prompt_slug`	No	Filter by prompt slug
`status_code`	No	Legacy Portkey query param for HTTP status codes. Comma-separated string; prefer status_codes for structured inputs.
`ai_org_model`	No	Legacy Portkey query param for provider/model pairs. Format: 'provider__model' with double underscore, e.g. 'openai__gpt-4' or 'anthropic__claude-3-opus'. Comma-separated string; prefer provider_models for structured inputs.
`config_slugs`	No	Structured alias for configs. Use an array of config slugs; normalized to the legacy comma-separated Portkey query param.
`status_codes`	No	Structured alias for status_code. Use an array of HTTP status codes; normalized to the legacy comma-separated Portkey query param.
`virtual_keys`	No	Legacy Portkey query param for virtual key slugs. Comma-separated string; prefer virtual_key_slugs for structured inputs.
`workspace_slug`	No	Filter by specific workspace
`metadata_filter`	No	Structured alias for metadata. Use an object such as { env: 'prod' }; normalized to a JSON string before the request is sent.
`provider_models`	No	Structured alias for ai_org_model. Use provider__model strings in an array; normalized to the legacy comma-separated Portkey query param.
`total_units_max`	No	Maximum number of total tokens to filter by
`total_units_min`	No	Minimum number of total tokens to filter by
`prompt_token_max`	No	Maximum number of prompt tokens
`prompt_token_min`	No	Minimum number of prompt tokens
`virtual_key_slugs`	No	Structured alias for virtual_keys. Use an array of virtual key slugs; normalized to the legacy comma-separated Portkey query param.
`completion_token_max`	No	Maximum number of completion tokens
`completion_token_min`	No	Minimum number of completion tokens
`weighted_feedback_max`	No	Maximum weighted feedback score (-10 to 10)
`weighted_feedback_min`	No	Minimum weighted feedback score (-10 to 10)
`time_of_generation_max`	Yes	End time for the analytics period (ISO8601 format, e.g., '2024-02-01T00:00:00Z')
`time_of_generation_min`	Yes	Start time for the analytics period (ISO8601 format, e.g., '2024-01-01T00:00:00Z')

Output Schema

TableJSON Schema

Name	Required	Description
`ok`	Yes	Whether the tool call succeeded and returned structured data
`data`	No	Structured success payload when ok is true
`error`	No	Structured error payload when ok is false

Schema Changelog

Changes observed during successful MCP inspections.

Changed16 schema fields changedv0.11.5Sep 8, 2026
- removedInput schema / properties / completion_token_max / exclusiveMinimum
```
Removed value: -0
```
- addedInput schema / properties / completion_token_max / minimum
```
Added value: +0
```
- removedInput schema / properties / completion_token_min / exclusiveMinimum
```
Removed value: -0
```
- addedInput schema / properties / completion_token_min / minimum
```
Added value: +0
```
- removedInput schema / properties / cost_max / exclusiveMinimum
```
Removed value: -0
```
- addedInput schema / properties / cost_max / minimum
```
Added value: +0
```
- removedInput schema / properties / cost_min / exclusiveMinimum
```
Removed value: -0
```
- addedInput schema / properties / cost_min / minimum
```
Added value: +0
```
- removedInput schema / properties / prompt_token_max / exclusiveMinimum
```
Removed value: -0
```
- addedInput schema / properties / prompt_token_max / minimum
```
Added value: +0
```
- removedInput schema / properties / prompt_token_min / exclusiveMinimum
```
Removed value: -0
```
- addedInput schema / properties / prompt_token_min / minimum
```
Added value: +0
```
- removedInput schema / properties / total_units_max / exclusiveMinimum
```
Removed value: -0
```
- addedInput schema / properties / total_units_max / minimum
```
Added value: +0
```
- removedInput schema / properties / total_units_min / exclusiveMinimum
```
Removed value: -0
```
- addedInput schema / properties / total_units_min / minimum
```
Added value: +0
```

Changed1 schema field changedv1.0.2Jul 4, 2026

changedInput schema / required

Previous value: -[
-  "time_of_generation_min",
-  "time_of_generation_max",
-  "api_key_ids"
-]New value: +[
+  "time_of_generation_min",
+  "time_of_generation_max"
+]

Addedv1.0.1May 7, 2026
RemovedMay 1, 2026
First observedJan 1, 2026

TDQS

A4.4/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already establish read-only, idempotent, non-destructive behavior, so the description's job is lighter. It adds meaningful context beyond annotations by disclosing the Enterprise licensing gate and the 403 error behavior, plus clarifying that the data is cache-hit-only and includes per-bucket time-series breakdown. This is useful behavioral information not present in the structured fields.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Three sentences: the first states what the tool returns, the second gives usage guidance and the alternative, and the third discloses access restrictions. Every sentence earns its place, and the core purpose is front-loaded.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the large sibling workspace, the description fully identifies this tool's niche among many analytics tools and provides the essential usage and failure conditions. The schema covers all 29 parameters and an output schema exists, so the description doesn't need to explain return structure. Minor gaps like pagination or bucket interval specification are not described, but they are not critical for tool selection or initial invocation.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema covers 100% of parameters with descriptions, including legacy/structured aliases and normalization behavior, so the schema carries the full semantic burden. The description does not need to add parameter-level detail; the only implicit addition, that the data is time-series with buckets, aligns with the required time_of_generation_min/max parameters. Baseline 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb ('Get'), a precise resource ('cache-hit-only latency time-series data'), and the exact fields returned (summary.total_latency, summary.avg_latency, per-bucket total/avg latency). It also explicitly differentiates itself from the sibling get_latency_analytics by the 'cache-hit-only' scope, so an agent can tell them apart without inspecting schemas.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives explicit selection criteria: 'Use this to evaluate cached-response speed; use get_latency_analytics for all requests.' It also warns that the tool is Enterprise-gated and returns 403 on non-Enterprise plans, which tells the agent when it is likely to fail and why.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server