holonovel

Derived Knowledge

manage_knowledge

Destructive

Build, rebuild, and inspect derived semantic indexes and typed knowledge graphs over Novel sources; rank indexed items and read graph relations without modifying authoritative files.

Instructions

Manage derived, in-memory knowledge over Novel sources: a semantic index for advisory text ranking and a typed knowledge graph. index_build and graph_build are audited, replace the prior projection in one step, and do not persist a separate copy; they are safe to rebuild from the authoritative Novel file, which is never modified. The read actions never mutate state. Use when: building or rebuilding the index or graph, checking staleness, ranking indexed items, or reading relations and graph structure. Do NOT use when: searching a bound ruleset's index — use manage_ruleset (action: search); recording knowledge — use manage_corpus; recording a relationship — use manage_relationship.

Input Schema

Name	Required	Description
`limit`	No	Maximum candidates to return (index_search; default 5).
`query`	No	Query text to rank candidates against (index_search).
`action`	Yes	index_* actions operate the semantic index; graph_* actions operate the knowledge graph.
`item_id`	No	Indexed item id to filter relations (index_relations).
`node_id`	No	Graph node id (graph_neighbors).

Output Schema

Name	Required	Description	Default
`text`	No	Human-readable text envelope mirroring the structured result (REQ-001, REQ-548a).
`status`	Yes	Machine-readable result status: OK, NEED_INPUT, WARNING, PARTIAL, or an error category (REQ-002, REQ-548c).

Schema Changelog

Changes observed during successful MCP inspections.

Addedv1.0.4Sep 30, 2026

TDQS

A4.6/5.0

Behavior5/5

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

Annotations declare destructiveHint=true and readOnlyHint=false, which alone would make an agent wary; the description resolves that by explaining exactly what is replaced (the prior projection), what is not created (no persisted copy), and that the authoritative Novel file is never modified — so the 'destructive' act is a safe rebuild. It also segregates read actions as non-mutating, which the annotations' tool-level flags cannot express since they are a single set for eleven actions.

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

Conciseness4/5

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

Purpose, then behavioral guarantees, then positive and negative routing — a sensible order with no filler. The second sentence is long and compound ('audited, replace the prior projection in one step, and do not persist a separate copy'), which costs a little scannability.

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?

For an eleven-action tool with a rich enum and an existing output schema, the description covers the things the schema cannot: which action families mutate, that rebuilds are safe and reversible from source, and where to go for the three neighboring intents. Return shapes are correctly left to the output schema.

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% and each parameter already names its owning action (index_search, index_relations, graph_neighbors), so the schema carries the load. The description adds no format, default, or constraint detail beyond what is structured, which is the baseline-3 case.

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?

States a specific verb and resource — derived, in-memory knowledge over Novel sources — and immediately decomposes it into the two concrete artifacts it manages: a semantic index and a typed knowledge graph. The action families (index_*/graph_*) are named in prose, so an agent knows the tool's surface without opening the schema.

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?

Explicit 'Use when' list covers all four legitimate intents (build/rebuild, staleness check, ranking, relation/graph reads), and the 'Do NOT use when' block routes three distinct wrong turns to manage_ruleset, manage_corpus, and manage_relationship by name. Nothing about selection is left to inference.

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