lineage_record

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

Annotations indicate readOnly=false, openWorld=false, destructive=false. The description explains a write operation that refuses atoms without citations, stores relationships as edges, and has special handling for corrections and subject_id (guardian-consent seam). It adds substantial context beyond annotations about refusal behavior, edge persistence, and the high bar of person_inference claims. No contradictions. Slight deduction because it could be more explicit about immutability or side effects on linked atoms (e.g., 'supersedes' makes old ones non-current), but that's mentioned.

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

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

The description is appropriately detailed, with paragraph breaks separating the core atom definition, edge semantics, and subject_id seam. The requiredness of rationale/evidence is front-loaded after the metaphor. Although long (over 200 words), it's structured and every block adds necessary context. No fluff—the 'story of this willow' metaphor is a nice touch but doesn't obscure meaning.

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

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

Given 12 params, 0% schema coverage, and no output schema, the description does a strong job covering the non-obvious parameters and behavioral rules. It addresses edge direction, refusal, scope confinement, and consent seam—highly non-trivial aspects. It doesn't explain return values or errors, but with no output schema that's not expected. It also doesn't mention how to form the evidence references (PR/commit/file/session) but lists valid types inline. Answering 'what happens on failure' could be added, but overall it's quite complete.

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

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

Schema coverage is 0%, so description carries full burden. It explains rationale (WHY required), evidence (at least one citation required), and all three edge parameters (supersedes, derived_from, motivated_by) with specific meanings and directionality. subject_id is described in detail as the guardian-consent seam. Other params like id, title, app_id, origin, authority, tags are self-explanatory from names but description doesn't elaborate on origin/authority tags—this is a minor shortfall since they might need disambiguation. However, given the complexity, the description covers the critical semantic parameters well.

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

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

The description clearly states that this tool records a provenance atom containing WHY and evidence, and explains what the atom answers (origin, reason, prior state). It uses specific verbs and resources (record, provenance atom, lineage_edges) and distinguishes itself from sibling lineage tools (lineage_link, lineage_why, lineage_list) by defining its role as creating atoms with required rationale/evidence, versus linking or querying.

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

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

The description explicitly says when to use this tool (to record the story of an entity, answering where/why/before) and conditions: rationale and evidence are required, direction in edges is queried not stored twice, corrections re-record same id, confined to store_scope. It also gives examples of edge types (supersedes, derived_from, motivated_by) and clarifies semantics for each. However, it doesn't explicitly name sibling tools as alternatives, but the guidance is clear enough to select this over lineage_link/lineage_why/lineage_list—though it doesn't directly state 'use lineage_link for x instead'. Minor gap, but the when/why context is strong.

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