Skip to main content
Glama
LiLara-AI

ShadowGraph

Official

shadowgraph_search

Read-onlyIdempotent

Search stored decisions, attempts, and facts by content keywords, with optional filters for kind, status, project, and confidence to retrieve matching records.

Instructions

Find decisions, attempts, and facts whose declared content fields contain every query term. shadowgraph_retrieve adds one-hop neighbours, shadowgraph_recall fuses ranks over scoped memory, shadowgraph_context builds a working set, shadowgraph_traverse walks from an id. Reads only. Schema keys never match, and a filter is never a content match.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
kindNoRestrict results to decisions or to attempts. A structured filter, never a content match.
limitNoMaximum items to return, 1-1000; the default is 50 and completeness.limitSource reports which applied. Out of range is rejected, never silently clamped.
queryNoWhitespace-separated terms, matched case-insensitively as substrings against the declared content fields (title, goal, chosen, assumption, evidence, alternative, attempt solution, attempt result, attempt reason, environment). Every term must match something, so more terms narrow the result. Omit to filter without searching.
offsetNoItems to skip before this window. Ordering is total and deterministic, so paging cannot drop or duplicate an item.
statusNoReturn only decisions in this lifecycle state. A structured filter, so matching it is never counted as a content match.
projectNoRestrict to one project. Omitted, decisions and facts are searched across projects, while memories stay in the default project and all-null scope.
sourceClassNoReturn only records carrying this claimed origin class. A structured filter, never a content match.
minConfidenceNoReturn only decisions whose current confidence is at least this value, 0-1. A structured filter, never a content match.

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault
pageYesThe window actually applied to the matching items.
itemsYesThe items in this window, in deterministic order.
completenessYesDeclares exactly what this response left out, so a truncated result can never look complete.

Schema Changelog

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

  1. Changed9 schema fields changedv0.40.1
    • addedInput schema / properties / kind / description
      Added value: +"Restrict results to decisions or to attempts. A structured filter, never a content match."
    • changedInput schema / properties / limit / description
      Previous value: -"Page size. Omitted means a declared default, never silent truncation."New value: +"Maximum items to return, 1-1000; the default is 50 and completeness.limitSource reports which applied. Out of range is rejected, never silently clamped."
    • addedInput schema / properties / minConfidence / description
      Added value: +"Return only decisions whose current confidence is at least this value, 0-1. A structured filter, never a content match."
    • changedInput schema / properties / offset / description
      Previous value: -"Page offset."New value: +"Items to skip before this window. Ordering is total and deterministic, so paging cannot drop or duplicate an item."
    • addedInput schema / properties / project / description
      Added value: +"Restrict to one project. Omitted, decisions and facts are searched across projects, while memories stay in the default project and all-null scope."
    • addedInput schema / properties / query / description
      Added value: +"Whitespace-separated terms, matched case-insensitively as substrings against the declared content fields (title, goal, chosen, assumption, evidence, alternative, attempt solution, attempt result, attempt reason, environment). Every term must match something, so more terms narrow the result. Omit to filter without searching."
    • addedInput schema / properties / sourceClass / description
      Added value: +"Return only records carrying this claimed origin class. A structured filter, never a content match."
    • addedInput schema / properties / status / description
      Added value: +"Return only decisions in this lifecycle state. A structured filter, so matching it is never counted as a content match."
    • changedOutput schema / (root)
      Previous value: -nullNew value: +{
      +  "description": "Matching records with their pagination and completeness declaration.",
      +  "properties": {
      +    "completeness": {
      +      "description": "Declares exactly what this response left out, so a truncated result can never look complete.",
      +      "properties": {
      +        "complete": {
      +          "description": "True only when every matching item is present.",
      +          "type": "boolean"
      +        },
      +        "contentFields": {
      +          "description": "The declared content fields a query term could match.",
      +          "items": {
      +            "type": "string"
      +          },
      +          "type": "array"
      +        },
      +        "limitSource": {
      +          "description": "Whose choice bounded the result.",
      +          "enum": [
      +            "caller",
      +            "default"
      +          ],
      +          "type": "string"
      +        },
      +        "losslessItems": {
      +          "description": "Always true: each returned item is a full-fidelity record, never a summary or a truncated field.",
      +          "type": "boolean"
      +        },
      +        "omitted": {
      +          "description": "total minus returned.",
      +          "minimum": 0,
      +          "type": "integer"
      +        },
      +        "returned": {
      +          "description": "Items in this response.",
      +          "minimum": 0,
      +          "type": "integer"
      +        },
      +        "scope": {
      +          "description": "The project, query, and structured filters that produced this result, so it explains its own derivation.",
      +          "type": "object"
      +        },
      +        "total": {
      +          "description": "Items that matched in total.",
      +          "minimum": 0,
      +          "type": "integer"
      +        }
      +      },
      +      "required": [
      +        "scope",
      +        "returned",
      +        "total",
      +        "complete",
      +        "omitted",
      +        "losslessItems",
      +        "limitSource",
      +        "contentFields"
      +      ],
      +      "type": "object"
      +    },
      +    "items": {
      +      "description": "The items in this window, in deterministic order.",
      +      "items": {
      +        "description": "One hit and the evidence for why it matched.",
      +        "properties": {
      +          "filters": {
      +            "description": "The structured filters that were applied.",
      +            "type": "object"
      +          },
      +          "matched": {
      +            "description": "The content fields that actually matched, never schema keys.",
      +            "items": {
      +              "type": "string"
      +            },
      +            "type": "array"
      +          },
      +          "matchedBy": {
      +            "description": "content when query terms matched, filter when only structured filters applied, graph when the item was pulled in as a neighbour.",
      +            "enum": [
      +              "content",
      +              "filter",
      +              "graph"
      +            ],
      +            "type": "string"
      +          },
      +          "reason": {
      +            "description": "Human-readable explanation built from matched, so it can never claim a match it cannot name.",
      +            "type": "string"
      +          },
      +          "record": {
      +            "description": "One stored entity: a decision, attempt, memory, fact, or alternative. Fields vary by kind, and an entity imported from an older schema may carry fewer of them.",
      +            "properties": {
      +              "actor": {
      +                "anyOf": [
      +                  {
      +                    "type": "string"
      +                  },
      +                  {
      +                    "type": "null"
      +                  }
      +                ],
      +                "description": "Who performed the write."
      +              },
      +              "client": {
      +                "anyOf": [
      +                  {
      +                    "type": "string"
      +                  },
      +                  {
      +                    "type": "null"
      +                  }
      +                ],
      +                "description": "Which client performed the write."
      +              },
      +              "createdAt": {
      +                "anyOf": [
      +                  {
      +                    "type": "string"
      +                  },
      +                  {
      +                    "type": "null"
      +                  }
      +                ],
      +                "description": "ISO 8601 creation time."
      +              },
      +              "id": {
      +                "description": "Stable entity identifier.",
      +                "type": "string"
      +              },
      +              "kind": {
      +                "description": "Entity kind: decision, attempt, memory, fact, relation, review, or alternative.",
      +                "type": "string"
      +              },
      +              "project": {
      +                "anyOf": [
      +                  {
      +                    "type": "string"
      +                  },
      +                  {
      +                    "type": "null"
      +                  }
      +                ],
      +                "description": "Project namespace; records imported from a schema that predates projects may carry null."
      +              },
      +              "schemaVersion": {
      +                "description": "Storage schema version this entity was written under. A value above the build’s own version is preserved rather than downgraded.",
      +                "type": "integer"
      +              },
      +              "sessionId": {
      +                "anyOf": [
      +                  {
      +                    "type": "string"
      +                  },
      +                  {
      +                    "type": "null"
      +                  }
      +                ],
      +                "description": "Session identifier recorded with the write."
      +              },
      +              "sourceClass": {
      +                "anyOf": [
      +                  {
      +                    "type": "string"
      +                  },
      +                  {
      +                    "type": "null"
      +                  }
      +                ],
      +                "description": "Claimed origin class recorded with the write. A claim, never proof."
      +              },
      +              "sourceRaw": {
      +                "anyOf": [
      +                  {
      +                    "type": "string"
      +                  },
      +                  {
      +                    "type": "null"
      +                  }
      +                ],
      +                "description": "The original origin label when it differed from sourceClass. Audit only; never evidence."
      +              },
      +              "updatedAt": {
      +                "anyOf": [
      +                  {
      +                    "type": "string"
      +                  },
      +                  {
      +                    "type": "null"
      +                  }
      +                ],
      +                "description": "ISO 8601 time of the last change."
      +              }
      +            },
      +            "type": "object"
      +          },
      +          "score": {
      +            "description": "Relevance score; 0 when no query terms were supplied.",
      +            "type": "number"
      +          }
      +        },
      +        "required": [
      +          "record",
      +          "score",
      +          "matched",
      +          "reason",
      +          "matchedBy",
      +          "filters"
      +        ],
      +        "type": "object"
      +      },
      +      "type": "array"
      +    },
      +    "page": {
      +      "description": "The window actually applied to the matching items.",
      +      "properties": {
      +        "hasMore": {
      +          "description": "True when matching items exist beyond this window.",
      +          "type": "boolean"
      +        },
      +        "limit": {
      +          "description": "Window size applied, whether the caller set it or the default did.",
      +          "minimum": 1,
      +          "type": "integer"
      +        },
      +        "offset": {
      +          "description": "Index of the first item returned.",
      +          "minimum": 0,
      +          "type": "integer"
      +        },
      +        "total": {
      +          "description": "Matching items before the window was applied.",
      +          "minimum": 0,
      +          "type": "integer"
      +        }
      +      },
      +      "required": [
      +        "offset",
      +        "limit",
      +        "total",
      +        "hasMore"
      +      ],
      +      "type": "object"
      +    }
      +  },
      +  "required": [
      +    "items",
      +    "page",
      +    "completeness"
      +  ],
      +  "type": "object"
      +}
  2. First observedv0.40.0

TDQS

A4.7/5.0
Behavior4/5

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

Annotations already declare readOnlyHint, openWorldHint, idempotentHint, and destructiveHint. The description adds non-obvious behavioral context: 'Schema keys never match, and a filter is never a content match' — an important guardrail for interpreting parameters. It could go further by noting pagination behavior, but the schema covers that.

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?

Four short sentences, each carrying information. The purpose is front-loaded, sibling distinctions are compact, and the read-only and filter semantics are stated in one concise clause. No filler or repetition beyond a harmless 'Reads only' echo of the annotation.

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?

Given a full input schema, output schema, and read-only/idempotent annotations, the description provides the missing high-level context: what counts as a match, how filters behave, and how this tool differs from the nearest alternatives. It is sufficient for an agent to select and invoke the tool correctly.

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

Parameters4/5

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

Schema description coverage is 100%, so the baseline is 3. The description adds cross-cutting meaning beyond individual parameter descriptions, especially the rule that structured filters are never content matches and that all query terms must appear in declared content fields. This helps an agent avoid misusing kind, status, sourceClass, and minConfidence.

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 opens with a specific verb and resource: 'Find decisions, attempts, and facts whose declared content fields contain every query term.' It also distinguishes itself from siblings by naming shadowgraph_retrieve, shadowgraph_recall, shadowgraph_context, and shadowgraph_traverse with their respective behaviors.

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 four sibling tools: retrieve adds one-hop neighbours, recall fuses ranks over scoped memory, context builds a working set, and traverse walks from an id. It also states 'Reads only,' helping an agent choose it over mutating alternatives.

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

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