Find retiring AI models and outdated API SDKs in your code, with dates, replacements and guides.
Scored across 7 tools
The set follows a clear scan/lookup/list triad pattern (find_retired_models vs lookup_model vs list_model_retirements) that descriptions distinguish well. The only real overlap risk is check_dependencies vs get_sdk_versions, which both report SDK version/deprecation status for npm/PyPI packages, though one is manifest-wide and the other single-package.
Every tool uses a consistent snake_case verb_noun pattern (check_, find_, get_, list_, lookup_). The verbs are varied but chosen deliberately to reflect the operation type, with no style mixing.
Seven tools is well-scoped for a focused SDK/model migration-checking server. Each tool earns its place, covering scanning, single lookups, listing, and guide retrieval without redundancy.
The domain of SDK and model retirement detection is well covered: manifest scanning, file scanning, single-item lookup, calendar listing, and migration guides. The one asymmetry is that model retirements have a list tool (list_model_retirements) while SDK deprecations do not have an equivalent browsable list, a minor gap agents can work around.
Check a dependency manifest (package.json, requirements.txt, pyproject.toml or Pipfile) for third-party API SDKs that are major versions behind, deprecated or replaced, against the live npm registry and PyPI. Returns each SDK's current spec, latest version, release-notes link and a Versyka migration guide when one exists.
| Name | Required | Description | Default |
|---|---|---|---|
| all | No | Check every package, not just third-party API SDKs. | |
| manifest | Yes | The full text of the manifest file. | |
| file_name | No | The file name, e.g. requirements.txt. Helps detect the format. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnly, idempotent, non-destructive and openWorld, so the safety profile is covered. The description adds real behavioral context beyond that: it performs live lookups against the npm registry and PyPI (implying network dependency and latency) and specifies what is returned — current spec, latest version, release-notes link, and a Versyka migration guide when one exists. It stops short of stating rate limits, failure behavior on an unparseable manifest, or format-detection fallback.
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, front-loaded with the action and scope, then the return payload. Every clause carries information: accepted formats, what is flagged, the registries consulted, and the output fields. 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 read-only, single-required-parameter tool with no output schema, the description covers both the input side (formats) and the output side (spec, latest version, release note, migration guide). Minor gaps remain around result size limits and behavior when the manifest format cannot be detected or parsed.
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 all three parameters (manifest, file_name, all) are already documented in the schema, including the meaning of 'all'. The description adds the enumeration of supported manifest types, which is minor value, so the baseline of 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 (check) and resource (a dependency manifest) and enumerates exactly which formats are understood: package.json, requirements.txt, pyproject.toml, Pipfile. It also names the target of the check (third-party API SDKs that are behind, deprecated or replaced) and the data source (live npm registry and PyPI), which cleanly separates it from the model-oriented siblings like find_retired_models and lookup_model.
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 scope — you call this when you have a manifest to audit — but the description never states when to prefer it over siblings such as get_sdk_versions or get_migration_guide, nor any prerequisite or exclusion. It is adequate but leaves routing to inference.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Scan source and config files for AI model IDs that the vendor (OpenAI, Anthropic, Google Gemini, Mistral, Cohere) or hosting platform (Amazon Bedrock, Groq) has retired or scheduled for retirement. Exact matches only, including Bedrock, Vertex AI, LiteLLM and OpenRouter spellings. Returns each model's shutdown date, days left, the vendor's recommended replacement, the source page and the file:line locations.
| Name | Required | Description | Default |
|---|---|---|---|
| files | Yes | Files to scan, each with its repository-relative path and full text. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnly, idempotent, openWorld and non-destructive, so safety is covered. The description adds genuinely useful behavior beyond that: exact-match-only semantics (no fuzzy detection), the included spelling variants (Bedrock, Vertex AI, LiteLLM, OpenRouter), and the shape of what comes back.
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 zero filler; the scan scope and match semantics are front-loaded, and the return contents close the statement.
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 returns and does so (shutdown date, days left, recommended replacement, source page, file:line). Combined with annotations, an agent has everything needed to call it correctly.
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 single 'files' parameter is fully documented in the schema, so the description is not required to explain it. It adds no path/content syntax detail beyond the schema, which is the baseline-3 case.
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?
Specific verb (scan) plus resource (source and config files for retired AI model IDs), and it scopes the detection domain to named vendors and hosting platforms. It is clearly distinguishable from siblings like list_model_retirements (which lists) and lookup_model (which looks up one model).
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 scan-oriented wording, but the description never says when to choose this over siblings such as list_model_retirements or lookup_model, nor does it state prerequisites. The 'exact matches only' note is a scoping constraint rather than a when-to-use rule.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
One migration guide as Markdown: what changed, each step with before and after code, how to find every affected call, how to test, and the vendor sources it is based on.
| Name | Required | Description | Default |
|---|---|---|---|
| slug | Yes | The guide slug from list_migration_guides, e.g. openai-python-v0-to-v1. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnly, idempotent, non-destructive and openWorld, so safety is covered. The description adds genuinely useful context beyond them by disclosing the return format (Markdown) and the guide's internal structure, which matters because there is no output schema.
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 naming the resource and format immediately, with the content breakdown following. The comma-separated inventory is dense but every clause conveys a distinct return element.
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 describing the return value and does so thoroughly. It omits error behavior for an invalid or unknown slug, which is the only notable gap for this simple read tool.
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% and the single slug parameter is fully documented with length bounds, a format example, and its source tool. The description adds no further parameter detail, so the baseline of 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 and resource ('One migration guide as Markdown') and enumerates the returned contents, so an agent can distinguish it from list_migration_guides. It stops short of explicitly naming that sibling or contrasting the two, which keeps it at 4 rather than 5.
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 never says when to reach for this tool versus list_migration_guides, nor does it mention prerequisites such as needing a valid slug from the listing tool. Usage is only implied by the singular 'One migration guide' phrasing.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
The latest version of an npm or PyPI package, whether it is deprecated, its release-notes link, the API vendor it belongs to and a Versyka migration guide when one exists.
| Name | Required | Description | Default |
|---|---|---|---|
| package | Yes | Package name, e.g. openai, @anthropic-ai/sdk, boto3. | |
| ecosystem | Yes | Which registry: npm (JavaScript/TypeScript) or pypi (Python). |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover the safety profile (readOnlyHint, idempotentHint, destructiveHint=false, openWorldHint), which is the heavy lifting. The description adds valuable return-content context – that deprecation status, a release-notes link, vendor, and an optional migration guide come back – but says nothing about caching, rate limits, or behavior when a package is unknown.
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 with no padding. It is a field list rather than a sentence about the action, but every clause conveys a real return dimension.
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 describing the return shape, and it does name the key fields an agent would want. It doesn't specify version format or error behavior for unknown packages, but it is sufficient for this simple two-parameter lookup.
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%, with both parameters documented and examples given for 'package' and an enum for 'ecosystem'. The description adds no syntax or format detail beyond the schema, 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 the specific resource (latest version of an npm/PyPI package) and enumerates the accompanying fields: deprecation flag, release-notes link, API vendor, and migration guide. Clear enough to select, but it does not contrast itself with siblings like get_migration_guide or list_migration_guides, so an agent must infer the boundary.
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?
There is no explicit when-to-use statement, no prerequisites, and no named alternative. An agent can only infer that this is the lookup for version/deprecation status rather than migration guides or dependency checks.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Versyka's step-by-step SDK migration guides (OpenAI, Anthropic, Google Gemini, AWS, Stripe, Firebase, Slack, Twilio, Auth0, Pinecone), optionally filtered by package name.
| Name | Required | Description | Default |
|---|---|---|---|
| package | No | Only guides for this package, e.g. openai. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnly, idempotent, openWorld and non-destructive, so the safety profile is covered. The description adds the concrete provider catalog, which is genuine context, but says nothing about result counts, ordering, or pagination behavior.
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?
One sentence, front-loaded with the resource, with the filter condition trailing. The provider list is long but serves as useful domain signal rather than 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 one-parameter, read-only listing tool whose annotations already carry the safety profile, this is largely sufficient. The only real omission is any hint of the return shape or volume, which matters slightly since no output schema exists.
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% and the single parameter already documents its purpose and gives an example ('openai'). The description only restates that filtering by package name is optional, adding no syntax or format detail beyond the schema. 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?
Names a specific verb+resource (list migration guides) and enumerates the ten supported SDK providers, which lets an agent judge coverage without opening anything. It does not explicitly distinguish itself from the singular sibling get_migration_guide, though the plural/enumerating framing implies 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?
"Optionally filtered by package name" implies when the parameter is worth using, but there is no guidance on when to call this versus get_migration_guide or check_dependencies. Usage is only weakly implied.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
The AI model retirement calendar: model IDs with their shutdown dates, replacements and sources, soonest first. Filter by vendor or by how many days ahead to look.
| Name | Required | Description | Default |
|---|---|---|---|
| vendor | No | ||
| within_days | No | Only retirements within this many days from today. | |
| include_retired | No | Also list models that are already retired. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already establish read-only, idempotent, non-destructive, open-world behavior, so the bar is lower. The description still adds real context beyond them: the returned fields (model IDs, shutdown dates, replacements, sources) and the sort order, which the annotations do not convey. It stops short of disclosing default filter behavior or result freshness.
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 tightly packed sentences, front-loaded with the returned content before the filtering options. The 'calendar' framing is slightly decorative but costs little, and there is no redundancy.
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 return-value burden and does so by naming the fields and ordering. Combined with annotations covering the safety profile, an agent has enough to call it, though the defaults for unfiltered calls and include_retired remain unstated.
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 67%: within_days and include_retired are self-documented, while vendor has only an enum with no prose. The description adds a plain-language gloss for vendor and within_days ('how many days ahead to look') but says nothing about include_retired or the default when no filters are passed. Baseline 3 is appropriate at this coverage.
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 concrete resource and verb-like framing ('The AI model retirement calendar: model IDs with their shutdown dates, replacements and sources') and states the ordering ('soonest first'). It never distinguishes itself from the close sibling find_retired_models, which an agent could easily confuse it with.
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 mentions what can be filtered ('Filter by vendor or by how many days ahead to look'), which implies usage, but gives no when-to-use/when-not guidance and no routing against find_retired_models, get_migration_guide, or lookup_model.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Is this AI model ID retired or scheduled for retirement? Accepts plain IDs and gateway spellings (e.g. openai/gpt-4-0613, anthropic.claude-3-sonnet-20240229-v1:0, claude-3-opus@20240229). Returns the shutdown date, status, replacement and source, or that no retirement is listed.
| Name | Required | Description | Default |
|---|---|---|---|
| model | Yes | The model ID as written in code or config. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnly, idempotent, openWorld and non-destructive, so the safety profile is covered. The description goes further by disclosing accepted input families (plain IDs and gateway spellings such as openai/gpt-4-0613 or claude-3-opus@20240229) and the shape of the answer, including the negative case ('no retirement is listed').
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, front-loaded with the question the tool answers, then inputs, then return contents. No filler and every clause is informative.
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 return-value burden and does so: shutdown date, status, replacement, source, or an explicit 'none listed'. Combined with the input-format guidance, an agent has everything needed to call it correctly.
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% and the single parameter is already documented, so baseline is 3. The description adds value the schema lacks: concrete examples of the multiple naming conventions the parameter accepts, which is the main practical ambiguity for this tool.
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 and resource ('look up a model's retirement') and frames the exact question it answers (retired or scheduled for retirement). This clearly separates it from siblings like list_model_retirements and find_retired_models, which cover sets of models rather than one ID.
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 question form implies usage for a single model ID, and the mention of accepted ID spellings hints at when it applies. However, it never names an alternative or states when not to use it (e.g. prefer list_model_retirements for bulk listing), so routing 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.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
list_model_retirements1 field changedInput schema / properties / vendor / enumPrevious value: -[
- "OpenAI",
- "Anthropic",
- "Google Gemini",
- "Mistral",
- "Cohere"
-]New value: +[
+ "OpenAI",
+ "Anthropic",
+ "Google Gemini",
+ "Mistral",
+ "Cohere",
+ "Amazon Bedrock",
+ "Microsoft Azure",
+ "Groq",
+ "Together AI"
+]check_dependenciesfind_retired_modelsget_migration_guideget_sdk_versionslist_migration_guideslist_model_retirementslookup_modelDated, verified API breakages for coding agents, plus a local scan for them in your existing code.
Is this model or API still safe to call? Retirement and breaking change dates, with the fix.
AI model shutdown dates, replacements, and migration diffs. Free, no API key.
Check if an AI model is deprecated, retiring, or silently changed price or context window.
Add one secure layer between your agents and this server.