This server gives AI agents post-quantum migration arithmetic and taxonomy facts — sizes, fragments, reassembly safety, failure families, and benchmark scoring — without exposing repair mechanisms.
credential_size – compute total on-wire bytes and per-component breakdown for a KEM+sig credential (e.g., ML-KEM-768 + ML-DSA-65).
fragments – calculate how many frames an object splits into on a transport and whether fragmentation is mandatory.
reassembly_window – determine if a safe capacity cap exists (feasible and safe) given object size, memory budget, and concurrency; if none, get the max safe concurrency.
list_failure_families – retrieve all 39 post-quantum migration failure families with case counts and published analogues.
describe_family – get details on a specific failure family: invariants broken, which unrepaired designs fail it, and what they did (no repair info).
score_submission – score a PQC-MFB submission (map of case_id to bool) to get coverage, regressions, and zero-coverage families.
All tools are exposed via MCP (JSON-RPC 2.0 over stdio) and return domain errors as tool results so agents can self-correct. Detection only — no repair mechanisms are provided.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@pqc-migration-mcp credential size for ML-KEM-768 and ML-DSA-44"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Give your AI agent the post-quantum migration facts it keeps guessing at.
Six tools over MCP: credential sizes, fragment counts, the reassembly window, the 39-family failure taxonomy, and benchmark scoring. Ask Claude "will our ML-KEM-768 handshake fit in a BLE MTU?" and it computes the answer instead of estimating one.
📖 Full documentation, tutorial and conceptual guide: https://nickharris808.github.io/pqc-toolkit/
Agents are increasingly doing PQC migration work, and they are confidently wrong about exactly the things that matter: how big a credential actually is, how many fragments it becomes, and whether a safe reassembly cap exists at your concurrency. Those are arithmetic, not judgement — so hand the agent the arithmetic.
The protocol layer here is dependency-free. MCP is JSON-RPC 2.0 over line-delimited stdio, which is small enough to implement directly and keeps the install trivial.
Related MCP server: attestix
pip install git+https://github.com/nickharris808/pqc-migration-mcpThis pulls pqc-sizes and pqc-mfb from their repositories too. Not on PyPI
yet, so pip install pqc-migration-mcp does not work today.
# talk to it directly -- it is line-delimited JSON-RPC on stdio
echo '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}' | pqc-migration-mcpAdd to claude_desktop_config.json:
{
"mcpServers": {
"pqc-migration": {
"command": "pqc-migration-mcp"
}
}
}Restart Claude Desktop. The six tools appear under the connector.
Tool | Answers |
| How many bytes is a KEM+signature credential, component by component? |
| How many fragments on this transport — and is fragmentation now mandatory? |
| Does a safe capacity cap exist at all? If not, what concurrency would work? |
| All 39 failure families, with case counts and published analogues |
| What breaks in this family, in which designs, and what did each do? |
| Score a PQC-MFB submission: coverage, regressions, zero-coverage families |
The transport is line-delimited JSON — one complete object per line. Keep the
request on a single line; a request wrapped across two lines arrives as two
incomplete ones and comes back as two -32700 parse errors.
$ echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"reassembly_window","arguments":{"largest_legitimate_object":12000,"memory_budget":32768,"concurrency":3}}}' | pqc-migration-mcpThe server replies with one JSON object per line. Pretty-printed, the content
payload of that reply is:
{
"budget": 32768,
"ceiling": 10922,
"concurrency": 3,
"explanation": "EMPTY WINDOW: floor 12,000 B > ceiling 10,922 B (short by 1,078 B). No capacity cap is both feasible and safe. Raise the budget to at least 36,000 B, reduce concurrency to at most 2, or choose a smaller credential.",
"floor": 12000,
"is_empty": true,
"max_safe_concurrency": 2,
"recommended_cap": null
}The agent gets a verdict and the number that would fix it, so it can propose a concrete change rather than reporting a problem.
It exposes detection. It does not expose repairs.
An agent can learn that a design fails krack_retransmission and exactly what the
unrepaired design did. It cannot obtain the mechanism that closes it. That boundary is
deliberate: an MCP tool returning repairs would let any user enumerate the entire closed
set in an afternoon.
There is a test that calls describe_family for all 39 families plus every other
tool, concatenates the responses, and fails if repair_mechanism, repaired_detail or
repaired_held appears anywhere in the output.
Domain errors — an unknown algorithm, an unknown family — come back as a tool result
with isError: true and a message naming the valid options, so the agent can correct
itself. Only protocol faults become JSON-RPC errors (-32601 unknown method/tool,
-32602 bad arguments, -32700 unparseable line).
A malformed line does not kill the loop; the server replies with a parse error and keeps serving.
pip install -e ".[dev]" && pytest # 57 passedTests cover the protocol, every tool, the moat boundary, and the real stdio transport driven as a subprocess — including a check that stderr stays empty, since MCP clients read stdout as protocol and stray warnings confuse them.
Arithmetic, taxonomy lookup and scoring. No cryptography, no network, no telemetry. It does not inspect your implementation. A clean answer means your configuration is sound, not that your code enforces it.
pqc-sizes · pqc-mfb ·
pqc-guard-action · pqc-dos-embedded
Closing the 39 families is what the closed core does. Relevant subject matter is covered by a filed provisional patent application. For commercial use of the full envelope, open a GitHub Discussion or an issue on this repository.
What this proves. That the arithmetic and taxonomy an agent is reasoning with are correct: real credential sizes, real fragment counts, a real window verdict, and the real failure taxonomy.
What it does NOT prove.
Not that the agent used the answer. This supplies facts; it does not supervise what is done with them.
Not an inspection of your code. No tool here reads your implementation.
Not a repair channel. Every tool exposes detection only. A test calls
describe_family for all 39 families plus every other tool and fails if a
repair field appears anywhere in the output.
Errors. Domain problems come back as tool results with isError: true and a
message naming valid options, so an agent can self-correct. Only protocol faults
become JSON-RPC errors.
Eleven free tools for teams moving authenticated key exchange to post-quantum. They find and measure; they do not repair.
Tool | What it does | Where |
Sizes, fragment counts, and the two-sided reassembly window | source | |
The same arithmetic for Node and the browser | source | |
Fail the build when the window is empty | GitHub Action | |
169 lines of C: the failure on a real 64 KB device | source | |
Re-verify the bound on-device, no SMT solver | source | |
The same bound in Lean 4 — 0 | source | |
The gate in synthesizable RTL, 5 Yosys proofs | source | |
pqc-migration-mcp ← you are here | Six MCP tools for AI agents | source |
322 cases · 39 failure families · scorer | source | |
The benchmark as a dataset | HF | |
122 named formal results, 6 provers | HF | |
Try it in your browser, no install | HF Space |
New here? The end-to-end tutorial walks one realistic migration through all of them in about ten minutes: sizes -> window -> CI gate -> benchmark.
In a hurry? pqc-sizes tells you in five seconds whether your credential fragments and whether a safe cap exists. pqc-explorer does the same in a browser, with no install.
Closing the 39 failure families — downgrade binding, retransmission-safe installation, fragmentation transcripts, roaming forward secrecy, multi-link key separation, admission control, group-key binding — is a separate proprietary codebase. Relevant subject matter is covered by a filed provisional patent application.
That split is measured, not asserted: under a replicate noise control only 4 of 32 repair mechanisms are externally distinguishable, so publishing these detectors does not disclose the repairs.
For commercial licensing, open a GitHub Discussion or an issue on any of these repos.
Apache-2.0. See LICENSE and CONTRIBUTING.md.
Total on-wire bytes for a KEM + signature credential, with a per-component breakdown.
| Name | Required | Description | Default |
|---|---|---|---|
| kem | No | KEM name, e.g. ML-KEM-768 | ML-KEM-768 |
| sig | No | Signature name, e.g. ML-DSA-65 | ML-DSA-65 |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the burden. It mentions a 'per-component breakdown' but does not specify the output format, side effects, or constraints like required permissions. Minimal disclosure.
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 that efficiently conveys the core function. However, front-loading could be improved by adding an explicit verb. Still well-structured.
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?
Adequate for a simple tool with two optional parameters, but lacks details on the return value format (e.g., boolean? object?). Without an output schema, more context would help.
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 descriptions for both parameters. The description repeats the concept but adds no new meaning beyond what the schema provides. Baseline 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 clearly states the tool computes on-wire bytes for a credential with a breakdown, which distinguishes it from sibling tools like list_failure_families. However, the verb is implied rather than explicit (e.g., 'calculate').
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?
No guidance on when to use this tool, when not to, or alternatives. The sibling tools are unrelated, but the description does not help the agent decide context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Detail for one failure family: the invariants it breaks, the unrepaired designs that fail it, and what each did. Does not return repairs.
| Name | Required | Description | Default |
|---|---|---|---|
| family | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations exist. Description mentions what is returned and what is not (repairs), but lacks information on side effects, permissions, or whether it is a read-only operation. Basic disclosure but not comprehensive.
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?
Single sentence, efficient and front-loaded with purpose. No redundant words, but a structured list of what is included might improve clarity without expanding length significantly.
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?
Describes output content (invariants, designs) but not structure or format. No output schema. Lacks guidance on the parameter value. Adequate for narrow use but insufficient for full autonomy.
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% for the only parameter 'family'. Description does not explain what the parameter value should be (e.g., family ID or name) or provide format examples. Fails to add meaning beyond the schema's type and required status.
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?
Description clearly states the tool provides detailed information for one failure family, including invariants and unrepaired designs, and explicitly excludes repairs. This distinguishes it from sibling tool list_failure_families, which likely lists all families.
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?
Implies use when details on a specific family are needed, but does not explicitly state when to use versus siblings like list_failure_families or other tools. No alternatives or exclusions provided.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
How many fragments an object becomes on a transport, and whether fragmentation is therefore mandatory.
| Name | Required | Description | Default |
|---|---|---|---|
| object_bytes | Yes | ||
| frame_payload | Yes | usable payload bytes per frame |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description must carry the full burden of behavioral disclosure. It indicates the tool calculates fragment count and mandatory status, but it does not disclose side effects, authorization needs, error conditions, or whether the operation is read-only. For a computation tool, the lack of safety information is a gap.
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 a single, concise sentence that immediately conveys the tool's purpose with no extraneous words. It is well-structured and front-loaded.
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 2-parameter tool, the description tells what the tool computes, but it lacks information about the return format (the output schema is absent). The agent must infer whether the result is a number, boolean, or structured object. This is a moderate completeness gap.
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% (frame_payload has a description). The tool description adds context by relating the parameters to object transport, but it does not explain what object_bytes is or provide details beyond the schema. It fails to compensate for the missing schema description of object_bytes.
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 the tool computes 'how many fragments an object becomes on a transport' and determines if fragmentation is mandatory. This is a specific verb+resource that distinguishes it from sibling tools like list_failure_families and reassembly_window.
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?
No guidance is given on when to use this tool versus alternatives. The description does not mention prerequisites, exclusions, or comparisons with sibling tools. The agent must guess the appropriate context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
All 39 post-quantum migration failure families, with case counts and published prior-art analogues.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries full burden. It mentions output content but doesn't disclose behavioral traits such as read-only nature, permissions needed, rate limits, or any side effects. For a tool with no annotations, this is insufficient.
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, clear sentence with no extraneous information. Every word adds value.
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 no output schema, the description provides reasonable context about return values (case counts, analogues). However, it lacks details like ordering, filtering, or any prerequisites. With no annotations, additional behavioral context would improve completeness.
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?
There are 0 parameters, so the schema provides no information. The description adds meaning by explaining what the tool returns, which is the full list. Baseline for 0 params is 4.
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 specifies the verb 'list' and resource 'failure families', explicitly states 'All 39', and includes details on return content (case counts and prior-art analogues). This distinguishes it from sibling tools like 'describe_family' which likely focuses on one family.
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 implies use case: get a comprehensive list of all failure families. It doesn't explicitly state when not to use or name alternatives, but the contrast with 'describe_family' is clear. No explicit exclusions or when-not guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
The two-sided reassembly-capacity window. Returns is_empty=true when NO capacity cap is both feasible and safe, plus the maximum concurrency that would be safe.
| Name | Required | Description | Default |
|---|---|---|---|
| concurrency | Yes | ||
| memory_budget | Yes | ||
| largest_legitimate_object | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, and the description does not fully disclose behavior. It lacks information on side effects, authentication, safety, or what 'feasible and safe' means. The description is insufficient for an agent to understand the tool's full 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?
The description is brief with one sentence, but it could be more structured. It front-loads jargon and then specifies returns. No superfluous words, but clarity is sacrificed for brevity.
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 only partially describes the return value (is_empty and max concurrency). It does not cover error conditions, edge cases, or other potential return fields. The description is incomplete for effective use.
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?
The input schema has no descriptions, and the tool's description does not explain the meaning of each parameter ('largest_legitimate_object', 'memory_budget', 'concurrency'). Minimal context is provided, leaving the agent guessing.
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 basic idea of the tool's purpose (computing a capacity window), but uses jargon ('two-sided reassembly-capacity window') and doesn't clearly state the action (e.g., 'compute' or 'get'). The return values are specified, providing some clarity.
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?
No guidance on when to use this tool versus its siblings. The description does not mention context, prerequisites, or alternatives.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Score a PQC-MFB submission ({case_id: bool}). Returns coverage, regressions, and which families have zero coverage.
| Name | Required | Description | Default |
|---|---|---|---|
| submission | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description discloses return values (coverage, regressions, zero-coverage families) but does not mention side effects, required authentication, or whether the operation is read-only. Since no annotations are provided, the description bears full burden, and the lack of side-effect clarity is a gap.
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 a single concise sentence that front-loads the verb and resource. However, the notation '{case_id: bool}' is somewhat cryptic and could be integrated into the schema or clarified.
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 the lack of output schema and detailed input schema, the description should provide more context on the input object structure and the exact format of the return values. It covers outputs but omits input details, making it incomplete for proper use.
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?
The input schema has 0% description coverage, and the description only hints at a 'case_id' field via '{case_id: bool}', which is not defined in the schema. The structure of the required 'submission' object is left entirely unexplained, so the description adds minimal value beyond 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 description clearly states the action (score) and the specific resource (PQC-MFB submission), and lists the outputs (coverage, regressions, zero-coverage families). This distinguishes it from sibling tools like list_failure_families or describe_family, which serve different purposes.
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 implies the tool is used when you need to evaluate a submission, but it does not provide explicit guidance on when to use it vs. siblings, nor does it mention prerequisites or avoidance scenarios.
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.
v0.1.0credential_sizedescribe_familyfragmentslist_failure_familiesreassembly_windowscore_submissionScored across 6 tools
Each tool targets a distinct aspect of PQC migration analysis: failure families, reassembly capacity, submission scoring, credential size, family details, and fragmentation. No overlaps in functionality.
Most tools follow a verb_noun pattern with underscores (list_failure_families, score_submission, describe_family). 'credential_size' and 'reassembly_window' are noun-like but still clear; 'fragments' is a single noun, slightly deviating.
The set includes 6 tools, which is well within the ideal 3-15 range. Each tool addresses a specific need without redundancy, making the scope manageable and focused.
The tools cover querying failure families and scoring submissions, but lack submission management, repair retrieval (noted in describe_family), and listing submissions. Some gaps exist for a full workflow.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Production-grade cryptography toolkit with 31 MCP tools for classical, PQC, and KMS workflows.
QuantumOracle — 18 post-quantum crypto tools: Kyber, Dilithium, hybrid schemes, migration.
AI-security knowledge as MCP: standards-mapped tools (OWASP, NIST, MITRE) for AI agents.
Tamper-evident proof creation and verification for AI agents via MCP, A2A, and REST.