Tracks Claude Code usage-limit resets announced by @ClaudeDevs and Claude Code team members, plus Codex resets from codex-resets.com. Not affiliated with Anthropic. Records announced resets and does not predict the next one. Read-only, no API key. Tools: get_last_reset, list_resets, get_reset_stats.
Scored across 3 tools
Each tool has a clearly distinct purpose: get_last_reset returns the most recent reset details, get_reset_stats provides aggregate statistics, and list_resets enumerates all tracked announcements. While get_last_reset includes some extra data (banked resets, last covering all users), its primary focus is unambiguous and does not overlap enough to cause misselection.
All tool names follow a consistent verb_noun snake_case pattern: get_last_reset, get_reset_stats, list_resets. The verbs 'get' and 'list' are used appropriately for their respective operations, with no stylistic deviations.
Three tools are well-scoped for this read-only tracking server: one for the latest reset, one for statistics, and one for listing announcements. Each tool earns its place without redundancy or bloat.
The surface covers the core needs—latest reset, list, and stats—but lacks a way to retrieve detailed information (e.g., plans covered, announcement date) for a specific non-latest reset. Agents can work around this by using list_resets for summaries, but a minor gap remains.
The most recent usage-limit reset: when it was announced, how many days ago, whether it was a banked reset that users apply themselves (and the deadline), which plans it covered, whether it has been reviewed, and the original post. If the latest reset did not cover every user, also returns the last one that did, and lists every banked reset that can still be applied. Use it for 'when did Claude last reset usage limits?' or 'do I have a banked reset to use?'.
| Name | Required | Description | Default |
|---|---|---|---|
| provider | No | claude (Claude Code, the default) or codex (OpenAI Codex). | claude |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already establish the safe, read-only, idempotent profile, and the description adds substantial return-behavior detail: it explains the fallback when the latest reset did not cover every user and lists every applicable banked reset. This is meaningful behavioral context beyond what the annotations provide.
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 front-loaded with the core object and its contents, followed by a compact use-case clause. Both sentences earn their place and contain no filler.
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 simple read-only tool with one schema-documented parameter and no output schema, the description supplies the necessary return semantics, fallback behavior, and usage examples. Annotations cover the safety profile, so nothing critical is missing.
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 the only parameter is a provider enum already well documented in the schema. The description mentions Claude in examples but adds no meaning beyond the schema's enum descriptions, so the baseline of 3 is appropriate.
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 and resource ('most recent usage-limit reset') and enumerates the returned fields, making the tool's scope clear. It does not explicitly distinguish itself from its siblings get_reset_stats or list_resets, so sibling differentiation is only implicit.
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 provides clear usage context through example questions: 'when did Claude last reset usage limits?' and 'do I have a banked reset to use?'. It does not state when not to use this tool or name the alternative sibling tools for other scenarios.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Counts and gaps for resets, as on claude-resets.com: resets announced, average, median and longest gap in days, days since the last, and resets per month. With provider 'both' and no date, the window is the period in which both are tracked, the fair comparison. Limit changes never count.
| Name | Required | Description | Default |
|---|---|---|---|
| since | No | Start the window on this UTC date. | |
| provider | No | both |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover the safety profile (readOnly, idempotent, non-destructive, closed-world), so the bar is lowered. The description adds real beyond-annotation substance: the exact set of returned metrics, the default window semantics for provider='both', and the data caveat 'Limit changes never count.'
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 sentences, metric list front-loaded, no filler. The phrasing is dense but every clause (metric enumeration, default-window rule, limit-change caveat) carries information.
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 correctly enumerates the returned statistics, and it discloses the default window and a data caveat. Only minor gaps remain (e.g., distinct meaning of 'claude' vs 'codex' windows), which are reasonably inferable.
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 only 50% ('since' is documented, 'provider' is not), so the description must compensate. It does: it explains that provider='both' without a date uses the overlap window, which is the meaningful semantic behind that enum value and its default interaction with 'since'.
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 names the resource (resets) and enumerates exactly what is computed: counts, average/median/longest gap, days since last, resets per month. This is clearly an aggregate-statistics tool, implicitly distinct from get_last_reset and list_resets, but the siblings are never named so differentiation is left to inference.
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 tells the agent what happens with the default (provider 'both' and no date selects the common tracked window, called out as 'the fair comparison'), which is useful context for choosing inputs. However, there is no explicit when-to-use-this-vs-get_last_reset-or-list_resets guidance or any exclusion.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tracked announcements, newest first: resets, limit changes (which raise or lower limits without resetting usage, and never count as resets), or both, for Claude, Codex or both, optionally only since a date.
| Name | Required | Description | Default |
|---|---|---|---|
| kind | No | reset | |
| limit | No | ||
| since | No | Only announcements on or after this UTC date. | |
| provider | No | claude |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare this is a read-only, idempotent, non-destructive, closed-world operation, so the burden is lower. The description does add real value by disclosing ordering (newest first) and the semantic rule that limit changes 'never count as resets,' but it says nothing about return format, truncation, or how the limit cap behaves.
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 sentence, front-loaded with the resource and ordering before the qualifiers. The nested parenthetical about limit changes is dense but earns its place by defining a subtle distinction; only minor restructuring would improve readability.
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 four optional parameters, no output schema, and annotations covering the safety profile, the description does an adequate job on filtering and ordering. It omits any explanation of the result page size cap, and gives no route to distinguishing this tool from its two siblings, leaving those gaps for the agent to guess.
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 only 25% (the kind and provider enums are undocumented in the schema), and the description compensates by spelling out the kinds (resets, limit changes, or both), the providers (Claude, Codex, or both), and the since-date filter. The one gap is the 'limit' parameter, whose cap and default behavior are never explained in either place.
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 resource (tracked announcements, listed newest first) and enumerates the two record types it returns (resets vs. limit changes), which is genuinely clarifying. It does not, however, differentiate itself from the siblings get_last_reset or get_reset_stats, so an agent still has to infer which listing tool to pick.
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 only usage hint is the trailing 'optionally only since a date,' which describes a parameter rather than when to choose this tool. Nothing says when to prefer this over get_last_reset or get_reset_stats, and there are no exclusions or prerequisites.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
get_last_resetget_reset_statslist_resets
Add one secure layer between your agents and this server.