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

Annotations only declare the safety profile (write, non-idempotent, non-destructive), while the description supplies real lifecycle semantics: the record starts unconfirmed, never counts toward proposals or debits, is invisible to aegis_scan, and only upgrades to 'W2' when the quote matches a hook-captured prompt in the same session. That is substantial post-write behavior disclosure, though the undefined 'W2' term and the absence of any return/error behavior keep it short of a 5.

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?

Two dense sentences with the core action front-loaded and the state-transition rules packed into the second. Efficient, but the second sentence is jargon-heavy and slightly overloaded for the value it conveys.

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?

For a two-parameter write tool with no output schema, the description covers what the call does and what happens to the record afterward, which is the critical context. Missing are the return value and any failure/validation behavior, but annotations plus schema carry the rest.

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 description coverage is 100%, so both 'quote' and 'paths' are already documented in the schema, and the description adds only the notion that the quote is a user correction. It offers no extra guidance on the optional 'paths' scoping behavior, so the baseline 3 applies.

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?

States a specific verb ('Record') and resource ('a user correction as an unconfirmed suggestion in user scope'), and explicitly distinguishes itself from the aegis_scan sibling by noting suggestions 'never appear in aegis_scan'. The core action is unambiguous, though the domain jargon (W2, proposals, debits) is left undefined.

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?

Usage is implied rather than stated: the agent must infer it should call this when the user issues a correction that should be captured but not yet counted. It clarifies one exclusion relative to aegis_scan, but never names when to prefer this over aegis_create_rule or aegis_report_outcome, nor what conditions make a suggestion appropriate.

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

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

Annotations already declare readOnlyHint=true, idempotentHint=true and destructiveHint=false; the description reinforces this with 'It writes nothing', which resolves the tension between the name 'create_rule' and a read-only hint. It also adds genuinely new context: the output is a prompt, not a saved rule, so the caller must persist it. It does not cover return format details beyond that.

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?

Two short sentences, zero filler, with the core transformation stated first and the critical caveat (nothing is written) immediately after. Every clause earns its place.

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?

With no output schema, the description carries the burden of explaining the return value, and it does so by saying the result is a prompt for drafting a rule rather than a rule itself. Combined with the manual-save instruction, an agent has enough to call and use it, though the prompt's shape is not further described.

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 description coverage is 100% and all three parameters (description, domain, rule_type with its enum) are documented in the schema, so the description adds no parameter-level meaning. Baseline 3 applies when the schema does the heavy lifting.

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 states a specific transformation (natural language concern -> prompt that drafts a governance rule), which is far more precise than the name 'create_rule' alone would suggest. It does not, however, name or contrast itself against any sibling such as aegis_list_rules or aegis_scan.

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?

It implies the usage context (drafting a rule before saving it manually) and gives a workflow instruction ('save the drafted rule under .aegis/rules/ yourself'), but never states when to prefer this tool over the siblings or any prerequisites. Usage is implied rather than spelled out.

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

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

Annotations already declare the safe read profile (readOnlyHint, idempotentHint, destructiveHint=false, openWorldHint=false), so the description's job is to add context beyond that. It does: unloadable files surface with an _error marker rather than failing, and stale_rules are defined relative to the last aegis sync. It does not disclose pagination or size limits, but for a local read this is solid added context.

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?

A single front-loaded sentence that leads with the tool's action and then unpacks the three output keys in matching order. It is dense with nested parentheticals, but nearly every clause carries information an agent needs to interpret the response.

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?

There is no output schema, so the description must carry the return-value burden, and it does: it defines all three top-level keys and the per-rule fields (type, hypothesis, ledger fitness). It omits ordering, pagination and response size, which keeps it short of a 5 for a tool with no output schema.

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 100% with both enum parameters fully described in the schema, so the baseline is 3. The description never mentions the view projection or rule_type filter, so it adds nothing to parameter understanding; the schema does all the work.

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?

Names a specific verb and resource ('list the governance rules') and immediately specifies the exact response shape ({rules, stale_rules, ignored}) with what each key contains. This is clearly distinguishable from siblings like aegis_create_rule or aegis_scan, which mutate or analyze rather than enumerate rules.

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?

Usage is implied by the verb and the described output (inspect rules, find stale ones, spot ignored markdown), but there is no explicit when-to-use statement, no mention of when a filter is preferable to a full listing, and no routing to or away from any sibling tool.

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

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

Annotations cover the safety profile (readOnlyHint=false, idempotentHint=true, destructiveHint=false) and the schema covers retry safety, yet the description adds genuinely non-obvious behavior: what actually gets credited, that rules_used is not the credit source, that unwitnessed claims earn nothing, and that reports are capped per session. The only gap is that the cap value and the returned payload are not characterized.

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 purpose is front-loaded in sentence one, which is good, but sentence two is a single run-on that packs crediting, tree matching, CI landing-commit requirements, unwitnessed-claim fallback and session caps into one unparsable clause chain. The content is dense with genuine information, yet the structure makes it hard to extract the actionable rule.

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?

For an 8-parameter mutation with no output schema, the description does cover the hard part — the crediting and capping semantics that an agent could never infer. However, it omits what the call returns, how the different outcome enum values affect crediting, and what the view projection yields, leaving the agent to guess at the result of the call.

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 description coverage is 63%, so the schema already documents most parameters (files, rules_used, idempotency_key, view, outcome). The description adds one meaningful clarification — that the rules_used array is not what is credited — but says nothing about outcome values, rework_count, tests_passed, notes, or the view projection, so it neither compensates for the uncovered third nor meaningfully exceeds the schema.

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 opening sentence gives a specific verb and object — report whether the rules you were shown helped — which is enough to place the tool as the feedback/outcome step of the aegis rule loop. It never names a sibling MCP tool (aegis_scan, aegis_create_rule) to sharpen the boundary, and the dense crediting explanation that follows muddies the one-line purpose rather than reinforcing it.

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?

It implies the required usage context (report only pays off when an `aegis check --record` ran against exactly the tree the server computes now) and states the fallback when that condition is absent. That is real when/when-not guidance, but it is framed as server-side crediting mechanics rather than as a rule for choosing this tool over its siblings.

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

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

Annotations already declare readOnly/idempotent/non-destructive, and the description goes well beyond them: it discloses token-budget clamping, the two-slot cap on mandatory rules, the existence of omitted_mandatory, that gates/invariants are served from the reviewed base commit and branch-local edits never bind (unreviewed_rules), staleness semantics, and that an unloadable rule file fails the whole scan with status FAIL. This is exactly the extra behavioral context the structured fields do not carry.

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 trigger condition is front-loaded and each subsequent sentence maps to a distinct output field or failure mode, so little is wasted. It is dense and slightly long, but the density is load-bearing given the absent output schema.

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?

With no output schema, the description must carry the return contract, and it does: it defines omitted_mandatory, rules_base, unreviewed_rules, stale_rules, and the FAIL behavior. An agent knows both how to invoke it and how to read the response.

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 100%, so both parameters and the enum values are already documented in the schema. The description references the file set only implicitly ('files you are about to edit'); it adds no format, ordering, or size guidance beyond what the schema states. Baseline 3 applies.

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 gives a specific verb plus resource ('returns the governance rules that apply to them') and a clear scope trigger ('Call before editing files'), which is enough to pick it over a plain rule lister. It never names a sibling tool explicitly (e.g. aegis_list_rules), so the differentiation from siblings is implied rather than stated.

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?

'Call before editing files' is an explicit, actionable when-to-use cue tied to the edit workflow. There is no when-not guidance and no named alternative for the opposite case (e.g. browsing all rules), so it stops short of full operative routing.

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