Skip to main content
Glama

agent-embassy

by io.github.tacticalnoot
OverviewDiscussionsAdmin

Compute Price Alerts

compute_price_alerts
Read-onlyIdempotent

FREE read-only price move alerts for the P2P GPU compute market. Scans consecutive snapshots in the Embassy market-intelligence log (vast.ai public asks) and flags per-GPU cheapest-price moves larger than 10%. Derived, not pushed: the alert log is the snapshot log; no webhook system. No payment required.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
gpuNoGPU class, e.g. H100 SXM, H100 NVL, H200, B200. Omit for all classes.
limitNomax alerts returned (default 50)

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault
alertsYes>10% cheapest-price moves, most recent first
marketYesMarket id: compute-spot
pairs_scannedYesConsecutive snapshot pairs compared
threshold_pctYesMove threshold in percent
history_sourceYesWhich market-intelligence log was scanned
snapshots_totalYesSnapshots scanned

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • changedOutput schema / (root)
      Previous value: -nullNew value: +{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "additionalProperties": false,
+  "properties": {
+    "alerts": {
+      "description": ">10% cheapest-price moves, most recent first",
+      "items": {
+        "additionalProperties": false,
+        "properties": {
+          "alert_at": {
+            "type": "string"
+          },
+          "direction": {
+            "enum": [
+              "up",
+              "down"
+            ],
+            "type": "string"
+          },
+          "from": {
+            "type": "number"
+          },
+          "from_snapshot_at": {
+            "type": "string"
+          },
+          "gpu": {
+            "type": "string"
+          },
+          "metric": {
+            "type": "string"
+          },
+          "pct_change": {
+            "type": "number"
+          },
+          "to": {
+            "type": "number"
+          }
+        },
+        "required": [
+          "alert_at",
+          "from_snapshot_at",
+          "gpu",
+          "metric",
+          "from",
+          "to",
+          "pct_change",
+          "direction"
+        ],
+        "type": "object"
+      },
+      "type": "array"
+    },
+    "history_source": {
+      "description": "Which market-intelligence log was scanned",
+      "enum": [
+        "live-log",
+        "image-seed"
+      ],
+      "type": "string"
+    },
+    "market": {
+      "description": "Market id: compute-spot",
+      "type": "string"
+    },
+    "pairs_scanned": {
+      "description": "Consecutive snapshot pairs compared",
+      "type": "number"
+    },
+    "snapshots_total": {
+      "description": "Snapshots scanned",
+      "type": "number"
+    },
+    "threshold_pct": {
+      "description": "Move threshold in percent",
+      "type": "number"
+    }
+  },
+  "required": [
+    "market",
+    "threshold_pct",
+    "history_source",
+    "snapshots_total",
+    "pairs_scanned",
+    "alerts"
+  ],
+  "type": "object"
+}
  2. Added

TDQS

A4.2/5.0
Behavior4/5

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

Annotations already declare readOnlyHint=true, idempotentHint=true, destructiveHint=false, so the safety profile is covered. The description adds valuable behavioral context: it scans consecutive snapshots, flags moves larger than 10%, is derived from the snapshot log, and has no webhook system. This goes beyond the annotations and clarifies the derived nature of the alerts.

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 with zero waste. The key facts are front-loaded: free, read-only, what it scans, what it flags, and the derived nature. Every sentence earns its place.

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?

The tool has an output schema, so return values are already documented. The description covers the input context (snapshot log, vast.ai public asks), the threshold, and the derived nature. It could mention pagination or the default limit, but the schema already covers the limit parameter. Overall, complete enough for an agent to call it correctly.

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?

Schema description coverage is 100%, so the schema already documents both parameters (gpu and limit). The description adds context for the 'gpu' parameter by giving examples (H100 SXM, H100 NVL, H200, B200) and clarifies that omitting it means all classes, but it doesn't add much beyond the schema. 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 states a specific verb ('scans', 'flags'), a precise resource (P2P GPU compute market price moves in the Embassy market-intelligence log), and a concrete threshold (>10% per-GPU cheapest-price moves). It clearly distinguishes itself from siblings by emphasizing it is derived and read-only, not a webhook system.

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

Usage Guidelines4/5

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

The description explicitly says 'FREE read-only' and 'Derived, not pushed: the alert log is the snapshot log; no webhook system', which tells the agent when to use this tool (to get computed alerts from snapshots) and what it is not (not a real-time push system). It does not explicitly name sibling alternatives like compute_history or compute_spot_prices, but the context makes the usage scenario clear.

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

All agent-embassy toolsInspect
Try in Browser
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources