Skip to main content
Glama
LiLara-AI

ShadowGraph

Official
by LiLara-AI
OverviewSchemaRelated ServersScoreDiscussions
JavaScript
Local

shadowgraph_maintain

Execute time-based maintenance to mark due decisions stale, expire facts, and evaluate reopen rules for accurate decision memory.

Instructions

Run time-based maintenance: make due decisions stale, expire due facts, then evaluate reopen rules. shadowgraph_review only evaluates reopen rules, shadowgraph_validate only reports, shadowgraph_update_status cannot set stale. Writes, and depends on the clock: a repeat at the same instant finds nothing to do but still commits a revision.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
nowNoISO 8601 instant to treat as the current time. Defaults to the real clock; supplying it makes the run deterministic.
factsNoFact key/value overrides evaluated instead of the stored facts of the same key. Stored facts are used for every key not listed here, so reopen rules still work after a restart.
changedFactsNoFact keys that just changed. Only string-form reopenWhen rules match this list; it is an ephemeral signal, not durable state.

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault
atYesThe instant maintenance ran against.
dueYesDecisions due for reconsideration after the run.
reviewSignalsYesEvery persisted review signal after the run, open and acknowledged alike.
agedDecisionIdsYesCompatibility alias of staleDecisionIds for older callers.
staleDecisionIdsYesDecisions moved to stale because they passed reviewAfter.

Schema Changelog

Changes observed during successful MCP inspections. Dates show when Glama detected each change.

  1. Changed5 schema fields changedv0.40.1
    • addedInput schema / properties / changedFacts / description
      Added value: +"Fact keys that just changed. Only string-form reopenWhen rules match this list; it is an ephemeral signal, not durable state."
    • addedInput schema / properties / changedFacts / items
      Added value: +{
+  "type": "string"
+}
    • addedInput schema / properties / facts / description
      Added value: +"Fact key/value overrides evaluated instead of the stored facts of the same key. Stored facts are used for every key not listed here, so reopen rules still work after a restart."
    • addedInput schema / properties / now / description
      Added value: +"ISO 8601 instant to treat as the current time. Defaults to the real clock; supplying it makes the run deterministic."
    • changedOutput schema / (root)
      Previous value: -nullNew value: +{
+  "description": "What maintenance changed and what is now due.",
+  "properties": {
+    "agedDecisionIds": {
+      "description": "Compatibility alias of staleDecisionIds for older callers.",
+      "items": {
+        "type": "string"
+      },
+      "type": "array"
+    },
+    "at": {
+      "description": "The instant maintenance ran against.",
+      "type": "string"
+    },
+    "due": {
+      "description": "Decisions due for reconsideration after the run.",
+      "items": {
+        "description": "One decision whose rejected alternatives are due for reconsideration.",
+        "properties": {
+          "alternativesToReconsider": {
+            "description": "Labels of the alternatives to look at again; all of them when the trigger was not alternative-specific.",
+            "items": {
+              "type": "string"
+            },
+            "type": "array"
+          },
+          "decisionId": {
+            "description": "The decision to reconsider.",
+            "type": "string"
+          },
+          "reason": {
+            "description": "Comma-separated causes: the fact keys whose reopenWhen rules matched, \"review date reached\", or \"decision outcome failed\".",
+            "type": "string"
+          },
+          "title": {
+            "anyOf": [
+              {
+                "type": "string"
+              },
+              {
+                "type": "null"
+              }
+            ],
+            "description": "Decision title, when the stored decision has one."
+          }
+        },
+        "required": [
+          "decisionId",
+          "reason",
+          "alternativesToReconsider"
+        ],
+        "type": "object"
+      },
+      "type": "array"
+    },
+    "reviewSignals": {
+      "description": "Every persisted review signal after the run, open and acknowledged alike.",
+      "items": {
+        "description": "A persisted review signal.",
+        "properties": {
+          "acknowledgedAt": {
+            "anyOf": [
+              {
+                "type": "string"
+              },
+              {
+                "type": "null"
+              }
+            ],
+            "description": "ISO 8601 time of the most recent acknowledgement."
+          },
+          "alternativesToReconsider": {
+            "description": "Alternative labels to look at again.",
+            "items": {
+              "type": "string"
+            },
+            "type": "array"
+          },
+          "createdAt": {
+            "description": "ISO 8601 time the signal was raised.",
+            "type": "string"
+          },
+          "decisionId": {
+            "description": "The decision this signal is about.",
+            "type": "string"
+          },
+          "id": {
+            "description": "Signal identifier, used by shadowgraph_ack_review.",
+            "type": "string"
+          },
+          "kind": {
+            "const": "review",
+            "description": "Always \"review\".",
+            "type": "string"
+          },
+          "reason": {
+            "description": "Why the signal was raised. Together with decisionId this is the dedupe identity, so the same cause never raises a second signal.",
+            "type": "string"
+          },
+          "status": {
+            "description": "open until acknowledged.",
+            "enum": [
+              "open",
+              "acknowledged"
+            ],
+            "type": "string"
+          },
+          "title": {
+            "anyOf": [
+              {
+                "type": "string"
+              },
+              {
+                "type": "null"
+              }
+            ],
+            "description": "Decision title, when the stored decision has one."
+          }
+        },
+        "required": [
+          "id",
+          "kind",
+          "decisionId",
+          "reason",
+          "alternativesToReconsider",
+          "status",
+          "createdAt"
+        ],
+        "type": "object"
+      },
+      "type": "array"
+    },
+    "staleDecisionIds": {
+      "description": "Decisions moved to stale because they passed reviewAfter.",
+      "items": {
+        "type": "string"
+      },
+      "type": "array"
+    }
+  },
+  "required": [
+    "at",
+    "staleDecisionIds",
+    "agedDecisionIds",
+    "reviewSignals",
+    "due"
+  ],
+  "type": "object"
+}
  2. First observedv0.40.0

TDQS

A4.7/5.0
Behavior5/5

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

Annotations already indicate this is not read-only and not idempotent, but the description adds crucial nuance: it commits a revision even when a repeat at the same instant finds nothing to do, and it depends on the clock. This goes beyond the structured metadata and helps an agent predict side effects.

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 carry all essential information with no repetition or filler. The core behavior is front-loaded, and the sibling comparisons are compact but informative.

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

Completeness5/5

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

With a complete input schema and an output schema present, the description needs only to explain behavior and tool selection, which it does thoroughly. The note about non-idempotence and clock dependence closes the most likely agent confusion.

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 has 100% coverage and already explains all three parameters in detail, including defaults and ephemeral semantics. The description does not add parameter-specific meaning beyond the schema, so the baseline of 3 applies.

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 action ('Run time-based maintenance') and breaks it into concrete steps: make decisions stale, expire due facts, evaluate reopen rules. It also names sibling tools and differentiates this one by what it does that they cannot. This leaves no ambiguity about the tool's role.

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 explicitly contrasts this tool with shadowgraph_review, shadowgraph_validate, and shadowgraph_update_status, stating what each alternative fails to do. It also clarifies that the tool writes and is clock-dependent, so an agent knows when to call it and what to expect.

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

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/LiLara-AI/shadowgraph'

If you have feedback or need assistance with the MCP directory API, please join our Discord server