Vaquill-AI/vaquill-mcp
OfficialRead-only MCP server for searching, resolving, and retrieving US primary law and related rulemaking/session-law data.
Search US statutes, regulations, constitutions, court rules, Federal Register documents, agency guidance, and other corpora.
Count sections by scope or Boolean expression, and extract/resolve statute citations from pasted text.
Resolve Bluebook citations to exact sections and fetch section metadata or full text.
Browse statutory hierarchies and inspect context: adjacent sections, definitions, citing sections, cross-state analogues, change history, and session-law enactments.
Explore federal rulemaking: Federal Register documents for CFR sections/parts, Regulations.gov dockets, RINs, Unified Agenda history, and Rules Committee records.
Work with state session laws: list, search, read records/text/versions, batch-fetch, and follow a change feed.
Check corpus coverage, live pricing, and credit balance.
Use generic
search/fetchtools for clients that require that compatibility pair.All tools are read-only; law-change alerts/boards/watches are not exposed as MCP tools.
This server has a remotely hosted endpoint. Open its connector listing for connection details, authentication requirements, and setup instructions:
vaquill-mcp
MCP server for Vaquill legal research API. Covers US federal and 50-state primary law: USC, CFR, state statutes and regulations, state and US constitutions, court rules, the Federal Register, executive orders, and agency guidance. Search primary law, resolve statutory citations, browse the hierarchy, and ground answers in official sources, all from your AI tools.
Quick Start
Prerequisites
Sign up at app.vaquill.ai/auth/signup to get your API key (new accounts get 1,000 free API credits, no card required). Steps for AI agents: vaquill.ai/agents.
Claude.ai (Web)
No installation needed. Add as a remote MCP server from Customize > Connectors > Add custom connector:
Option A: Bearer token (recommended)
Open the Request headers section of the same dialog and add the credential there:
URL: https://mcp.vaquill.ai/mcp
Header name: Authorization
Header value: Bearer vq_key_your_key_hereClaude sends the value exactly as you type it and adds no scheme of its own, so the value
field holds Bearer vq_key_... and not Authorization: Bearer vq_key_....
Option B: API key in the URL (compatibility)
https://mcp.vaquill.ai/s/vq_key_your_key_hereStill live, and still the only form that works where a client cannot send a header at all. Prefer Option A everywhere else: a credential in a URL is recorded in server logs, proxies and browsing history, which is why Anthropic's connector guidance advises against it and the MCP specification prohibits it.
Available on Claude Pro, Max, Team, and Enterprise plans. The Request headers section is in beta and is enabled per account, it accepts a short allowlist of header names, and it holds at most four. On Team and Enterprise an owner adds the connector under Organization settings first.
Claude Desktop
Open the config file from Settings > Developer > Edit Config:
macOS:
~/Library/Application Support/Claude/claude_desktop_config.jsonWindows:
%APPDATA%\Claude\claude_desktop_config.jsonLinux:
~/.config/Claude/claude_desktop_config.json
{
"mcpServers": {
"vaquill": {
"command": "uvx",
"args": ["vaquill-mcp"],
"env": {
"VAQUILL_API_KEY": "vq_key_your_key_here"
}
}
}
}For Indian legislation, add --jurisdiction IN. One process serves one
corpus, so run two entries if you want both:
{
"mcpServers": {
"vaquill": {
"command": "uvx",
"args": ["vaquill-mcp"],
"env": { "VAQUILL_API_KEY": "vq_key_your_key_here" }
},
"vaquill-india": {
"command": "uvx",
"args": ["vaquill-mcp", "--jurisdiction", "IN"],
"env": { "VAQUILL_API_KEY": "vq_key_your_key_here" }
}
}
}Quit Claude completely and reopen it. It does not reload the file.
Claude Code
Remote (no install):
claude mcp add --transport http --scope user vaquill \
https://mcp.vaquill.ai/mcp \
--header "Authorization: Bearer vq_key_your_key_here"Local (via uvx):
claude mcp add --scope user vaquill -e VAQUILL_API_KEY=vq_key_your_key_here \
-- uvx vaquill-mcp--scope user registers the server for every project; the default is the current directory
only. -e stores the key with the registration, so it survives Claude Code being launched
from an IDE, which an export in your shell profile does not. Both --header and -e are
variadic, so they have to come after the server name.
Cursor
Edit ~/.cursor/mcp.json for every project, or .cursor/mcp.json for one.
Remote:
{
"mcpServers": {
"vaquill": {
"url": "https://mcp.vaquill.ai/mcp",
"headers": {
"Authorization": "Bearer vq_key_your_key_here"
}
}
}
}Local (via uvx):
{
"mcpServers": {
"vaquill": {
"command": "uvx",
"args": ["vaquill-mcp"],
"env": {
"VAQUILL_API_KEY": "vq_key_your_key_here"
}
}
}
}VS Code (Copilot)
Add to .vscode/mcp.json. For every project instead of one, run the
MCP: Open User Configuration command and use the same shape there.
Remote:
{
"inputs": [
{
"type": "promptString",
"id": "vaquill-authorization",
"description": "Authorization header value",
"password": true
}
],
"servers": {
"vaquill": {
"type": "http",
"url": "https://mcp.vaquill.ai/mcp",
"headers": {
"Authorization": "${input:vaquill-authorization}"
}
}
}
}The credential goes in a prompt rather than in the file, so the file is safe to commit. VS Code asks for it on first start and remembers it. Restart the server after saving, or the tools will not appear.
Local (via uvx):
{
"servers": {
"vaquill": {
"type": "stdio",
"command": "uvx",
"args": ["vaquill-mcp"],
"env": {
"VAQUILL_API_KEY": "vq_key_your_key_here"
}
}
}
}Windsurf
Add to ~/.codeium/windsurf/mcp_config.json.
Remote:
{
"mcpServers": {
"vaquill": {
"serverUrl": "https://mcp.vaquill.ai/mcp",
"headers": {
"Authorization": "Bearer vq_key_your_key_here"
}
}
}
}Windsurf uses serverUrl for a remote server, not url, and interpolates ${env:VAR} if you
would rather read the credential from the environment.
Local (via uvx):
{
"mcpServers": {
"vaquill": {
"command": "uvx",
"args": ["vaquill-mcp"],
"env": {
"VAQUILL_API_KEY": "vq_key_your_key_here"
}
}
}
}Related MCP server: lawyer-mcp
Available Tools
Tools are generated from the live Vaquill API's OpenAPI spec at startup, so the
set always matches the current API. For the authoritative, up-to-date list and
per-call credit costs, run the free get_pricing tool or inspect your MCP
client's tool list.
The US server publishes 37 tools and the India server 22, counting search and
fetch. Every one is read-only. Law change alerts (boards and watches) are deliberately not MCP
tools: the API still serves them, so create and manage alerts in the Vaquill
console or over the REST API directly.
The main groups (representative tools shown):
US statutes & regulations
USC, CFR, all 50 state codes, constitutions, state court rules, the Federal Register, and agency guidance.
Tool | Description |
| Hybrid semantic + keyword search; filter by |
| Exact section count for a scope, or for a Boolean expression ( |
| Section metadata by |
| Full HTML + plain text of a section. |
| Metadata for up to 50 sections in one call. |
| Resolve a Bluebook citation (e.g. |
| Browse the statutory hierarchy one level at a time. |
| Self-describing coverage matrix: which corpora exist in which jurisdiction. |
Reading a section in context
Tool | Description |
| The sections immediately before and after, in statutory order. |
| The defined terms that govern a section, from its chapter's definitions section. |
| Which USC/CFR sections cross-reference this one (the inverse of |
| Provisions in other states addressing the same subject, ranked by similarity. |
| What our refreshes observed changing on this section over time. |
Federal rulemaking
The Federal Register lineage of the CFR, and the dockets and Unified Agenda entries that share a Regulation Identifier Number (RIN). Metadata only: no comment text. Costs are appended to each tool's description at startup from the live pricing table.
Tool | Description |
| The Federal Register documents behind one CFR section, from its printed source note and eCFR amendment dates, with a |
| Every Federal Register document that references a CFR title and part, newest first, over the most recent |
| One Regulations.gov docket: agency, type, title, RIN, document counts and comment count ( |
| Dockets, most recently modified first, filtered by |
| The documents of one docket, oldest first, with comment periods and Federal Register document numbers. Cursor paged. |
| Everything held for one RIN: Unified Agenda entry, OIRA reviews, dockets, Federal Register documents by role, CFR parts and related RINs. |
| The Unified Agenda edition history of one RIN, newest first, with |
| Judicial Conference Rules Committee records (minutes, agenda books, reports, rules packages, comments), most recent first, filtered by |
| One committee record with its comment dockets, Federal Rule links and a window of page text ( |
| What the rulemaking store holds per source, with a freshness verdict (free). |
Utility
Tool | Description |
| Live API credit pricing (free, no auth). |
| Generic one-string corpus search returning |
| Generic one-string retrieval returning |
search and fetch exist because ChatGPT's deep research and company-knowledge
connectors match a corpus server by those exact names and their single-string
signature, and refuse to work without them. They are thin wrappers over the
typed tools above, which any client that can call them should prefer: the typed
ones filter by jurisdiction, corpus, date and status, and these two do not.
fetch is deliberately lenient about its id, accepting an act_id, a citation
URL, a bare path, or a Bluebook citation such as 42 U.S.C. 1983.
Resources and prompts
This server is not only tools. An MCP server generated from an OpenAPI document publishes one tool per endpoint and nothing else; these two primitives are where the knowledge that is not in the API lives, and neither costs anything in the per-turn tool budget.
Resources (read, don't call):
Resource | Contents |
| How to use the corpus correctly: identifier rules, what "still good law" actually means here, what an empty change list does and does not tell you, and where the cost is. Not backed by any endpoint. |
| Coverage matrix: every corpusType and its per-jurisdiction counts. Free. |
| The filter vocabulary the India corpus actually holds. Free. |
| Live credit pricing. Free. |
Prompts (workflows, with the traps built in):
Prompt | What it encodes |
|
|
| Check coverage first, so "no such law" and "not in our corpus" are never conflated. |
| Change capture is observation, not effect; an empty result is not "nothing was amended". |
| Batch-resolve rather than looping, then verify the passage's claims against each section. |
| India: map IPC/CrPC to the BNS/BNSS provision in force since 1 July 2024 before answering. |
| India: read the filter vocabulary first, and state the amendment caveat correctly. |
Indian legislation
A separate endpoint on the same host. Central and State Acts plus the instruments of the principal regulators (SEBI, RBI, MCA, IRDAI, TRAI, DGFT): 22,265 enactments and 1,098,577 individually addressable sections.
https://mcp.vaquill.ai/in/mcp with Authorization: Bearer vq_key_your_key_here
https://mcp.vaquill.ai/in/s/vq_key_your_key_here key in the path, as aboveSame API key, same credit balance as the US endpoint.
Tool | Description |
| Search Indian legislation down to the section. |
| Browse and filter enactments by jurisdiction, regulator, year, status. |
| The filter vocabulary the corpus actually holds, with counts. |
| Source links (text, PDF, HTML) for one enactment. |
| Amendment history: what changed, by which Act, effective when. |
| The same generic pair described above, over Indian legislation. |
| Map a repealed criminal code to its 2023 replacement (IPC to BNS, CrPC to BNSS). |
One endpoint serves one jurisdiction, deliberately. The US and India
OpenAPI documents are disjoint, and each app derives its entire tool set from
one of them, so nothing under /mcp or /s/ can expose an Indian tool and
nothing under /in/ can expose a US one. The mount path selects an app; it
does not filter one, so there is no per-request check to get wrong. That also keeps an integrator's context window
to the jurisdiction they actually work in.
Indian regulatory publications
Circulars, orders, notices and press releases from named Indian regulators,
captured from the publishers themselves. A separate server on its own
origin, not a path on mcp.vaquill.ai:
https://reg-mcp.vaquill.ai/mcp Authorization: Bearer vq_key_..., or sign in over OAuth
https://reg-mcp.vaquill.ai/s/vq_key_... key in the path, for clients that cannot send a headerSame API key and credit balance as the other endpoints. The key needs the
single reg:read permission, which covers every tool and every publisher.
Tool | Description |
| Search by words and filters: publisher, source, document type, publication dates. Cursor paged. |
| One document by id, with the extracted text, the latest diff or every capture on request. |
| The archived original, returned as a file. |
| What is new, changed or retracted since a time or cursor, in capture order. |
| What the corpus covers and how fresh each source is. |
Capture began on 2026-08-23. An empty result for an earlier date does not mean nothing was published. Costs are appended to each description at startup from the live pricing table.
Why a subdomain. OAuth only works at an origin root. Discovery
(RFC 9728) is served from the origin's /.well-known/*, and a FastMCP app knows
only its own path, so an app mounted under a prefix advertises the wrong
resource URL and serves its document at the wrong location. Claude requires
resource to match the URL the user typed, path included. That is why the
India legislation endpoint above is bearer-only, and why this corpus is not
mounted there.
Running and deploying it
It is the same package and the same Dockerfile as mcp.vaquill.ai, with a
different entry point. It is deliberately not in remote_main._MOUNTS, which
drives the live US and India connectors.
pip install "vaquill-mcp[remote]"
PORT=8001 vaquill-mcp-regulatory # or: python -m vaquill_mcp.regulatory_main
curl localhost:8001/health # {"status":"ok",...,"corpus":"regulatory"}To deploy, create a second application from this repository and build the
existing Dockerfile with one build argument:
SERVER=vaquill-mcp-regulatoryLeaving it unset builds the vaquill-mcp-remote image exactly as before. Point
the new host name at the application and set its own environment (the table
under "Hosted server only" applies unchanged, with these differences):
VAQUILL_PUBLIC_URLis this origin,https://reg-mcp.vaquill.ai.With OAuth on, register
https://reg-mcp.vaquill.ai/auth/callbackas a second redirect URI on the upstream client. Give it its ownVAQUILL_OAUTH_JWT_SIGNING_KEYand its own Redis database or instance, so the two deployments do not share OAuth state.Without the
VAQUILL_OAUTH_*group it serves bearer keys only.HOSTandPORTbehave as above (0.0.0.0and8000).
At startup it reads the India pricing matrix and logs a warning naming any of the five tools that has no price row; such a tool is published without a cost line rather than with a guessed one.
Environment Variables
Variable | Required | Default | Description |
| Yes | - | API key ( |
| No |
| API base URL |
| No |
| Request timeout in seconds |
| No |
| stdio server only. |
Hosted server only
The variables below apply to vaquill-mcp-remote (the deployment behind
mcp.vaquill.ai) and are ignored by the stdio server.
OAuth is OFF unless the whole VAQUILL_OAUTH_* group is set, and the server
behaves exactly as it did before OAuth existed. A PARTIAL group raises at
startup rather than serving a half-built discovery document, because Claude
caches discovery globally by URL for about five minutes across all users, so a
wrong one outlives the misconfiguration that produced it.
Variable | Required | Default | Description |
| No |
| Listen address |
| No |
| Listen port |
| With OAuth | - | This server's public URL ( |
| With OAuth | - | Shared secret for resolving an OAuth user to their |
| With OAuth | - |
|
| With OAuth | - |
|
| With OAuth | - |
|
| With OAuth | - |
|
| With OAuth | - | The STATIC client registered upstream. Dynamic client registration stays off; this server is itself the DCR/CIMD facade. |
| With OAuth | - | Secret for that client |
| No |
| Must match the upstream signing key. |
| No |
| Expected |
| No | unset (any redirect accepted) | Comma-separated redirect URI patterns a client may register, e.g. |
The only redirect URI registered upstream is https://mcp.vaquill.ai/auth/callback.
Do NOT also register Claude's (https://claude.ai/api/mcp/auth_callback) or Claude
Code's loopback: this server proxies the flow, so the upstream never sees them, and
the ephemeral-port matching RFC 8252 requires is handled here.
Example Usage
Once configured, you can ask your AI assistant things like:
"What does 17 CFR 240.10b-5 say about insider trading?"
"Resolve 42 U.S.C. ยง 1983 to its section and show the full text"
"Find California statutes on tenant repair obligations"
"Browse the Texas statutory codes, then drill into the Penal Code"
Development
# Clone and install
git clone https://github.com/Vaquill-AI/vaquill-mcp.git
cd vaquill-mcp
uv sync --all-extras
# Run locally
VAQUILL_API_KEY=vq_key_... uv run vaquill-mcpTests
The suite needs no API key and no network access. Every test runs against the
checked-in OpenAPI fixtures in tests/fixtures/, and tests/conftest.py makes
any unmocked HTTP call raise, so a clone plus uv sync is the whole setup.
uv sync --extra dev --extra remote
uv run pytest # 407 tests
uv run pytest -W error::DeprecationWarning # what CI runsCI (.github/workflows/test.yml) runs two jobs, and they are deliberately not
redundant. test runs the suite against the locked environment on Python
3.10-3.13. wheel builds the package, installs it into a fresh venv without
the lockfile, and runs the same suite there. The second reproduces what a
customer gets from uvx vaquill-mcp: a lockfile does not constrain anyone
installing the published wheel, so a dependency shipping a breaking major shows
up in that job and nowhere else.
# Test with FastMCP inspector
uv run fastmcp dev src/vaquill_mcp/server.pySecurity
Report a vulnerability to contact@vaquill.ai (see SECURITY.md and https://www.vaquill.ai/security/vulnerability-disclosure).
Two things worth knowing before you connect this to an agent that holds other tools:
The server is read-only. Every tool carries
readOnlyHint: trueanddestructiveHint: false, and none writes, sends or deletes anything. The hints describe what the tools can do; they are not a security control, and a client should treat annotations from any server as untrusted unless it trusts the server.Returned text is third-party content. Statutes, notices, agency decisions and session laws come from government publishers (and, for some corpora, from vendor collection and OCR). We sanitise markup and restrict source links to government hosts, but we do not filter wording for instructions, and
openWorldHint: falsedescribes reach, not provenance. The server sends aninstructionsnote saying so; clients that ignore it should still treat tool output as untrusted.
How It Works
This package is a thin MCP wrapper around the Vaquill Developer API. At startup, it fetches the OpenAPI spec from the live API and auto-generates MCP tools using FastMCP. Tool names are derived automatically from each endpoint's OpenAPI operation id, so new API endpoints show up as clean, ready-to-use tools with no package update; key descriptions are refined for optimal LLM performance. The one exception is a short exclusion list in server.py (_ROUTE_MAPS): the SSE streaming route, and the law change alert routes, which are a product decision rather than a technical limit.
Because the spec is fetched at startup (not bundled), tools automatically reflect any API changes without a package update.
Credits & Pricing
API calls consume credits. The credit costs in the tables above are current at
the time of writing; the get_pricing tool and each tool's own description
always reflect the live price, so treat those as authoritative if they differ.
1 credit = $0.01 USD
License
MIT
Community
Questions, ideas, or want to contribute? Join the Vaquill community on Discord.
Available Tools
37 toolscount_statute_sectionsCount Sections in a ScopeARead-onlyInspect
Exact section count for a jurisdiction, corpus, code, title, chapter or part, or the number of sections matching a Boolean expression (query with matchType boolean, syntax as search_us_statutes). Use to size a job before walking divisions or paging a Boolean search. A natural-language question cannot be counted. isExact false means an upper bound (phrases, proximity, or filters beyond corpusType, state and status). A zero count may indicate a wrong filter.
Inputs: code, part, query, state, +8 more. Returns: object
extract_statute_citationsExtract Citations from TextARead-onlyInspect
Paste text (a brief, memo or answer) and get each distinct US statute, regulation, court rule and constitution citation, resolved to its section. Not case citations; extraction can miss one. Extraction is free (resolve=false); each distinct citation resolved is charged. Rows carry charStart/charEnd, rangeRole/rangePartner (range ends are two charged rows; interior rows are never charged) and counts.distinctSections. Up to 200,000 characters, never stored.
Inputs: text (required), state, resolve, Idempotency-Key, +1 more. Returns: object
fetchFetch DocumentARead-onlyInspect
Fetch the full text of one US law section by the id from a search result, returning {id, title, text, url, metadata}. Also accepts a citation URL, a bare path, or a Bluebook citation such as '42 U.S.C. 1983'. Present for clients that require the standard search/fetch pair; prefer get_us_statute_section_text if available. Charged as a section lookup plus a body read.
Inputs: id (required). Returns: object
get_credit_balanceCredit BalanceARead-onlyInspect
The credits this API key can spend now, calculated from live credit buckets under the same expiry rules used for billing. Use for pre-flight checks or low-balance alerts. creditsRemaining is a spendable balance, not a cached estimate. Any valid API key works without a particular scope, and normal rate limits still apply.
Inputs: no named parameters. Returns: object
get_part_rulemakingFederal Register Documents for a CFR PartARead-onlyInspect
Every Federal Register rule, proposed rule and notice whose CFR references include one title and part, newest first. The window is bounded to the years most recent publication years (default 5, max 12), and coverage says when a year was truncated. Use get_section_rulemaking for one section, and get_rulemaking_rin for the whole rulemaking behind a RIN.
Inputs: part (required), title (required), years, Idempotency-Key. Returns: object
get_pricingCredit PricingARead-onlyInspect
Current API credit pricing: per-endpoint credit costs and the credit-to-currency conversion rate (1 credit = $0.01 USD). Free, and no authentication required. Use to check what a call will cost before making it.
Inputs: region. Returns: object
get_rulemaking_committee_recordRules Committee RecordARead-onlyInspect
One Rules Committee record by recordId (rcr_ plus 24 hex): its comment dockets, linked Federal Rules (method, confidence) and a window of page text (pageFrom, pageLimit; nextPage continues). Drafting history, not law. quality needs_ocr: no text yet; quality blank: really empty. Comments and suggestions (textWithheld: true) withhold title, filename, URLs and text. An unknown record is a refunded 404.
Inputs: record_id (required), pageFrom, pageLimit, Idempotency-Key. Returns: object
get_rulemaking_coverageRulemaking CoverageARead-onlyInspect
What the rulemaking store holds, per source: row counts, newest change, last successful refresh and a freshness verdict (ok, stale or unmeasured), the Unified Agenda editions held, and what is not included and why. Read it before treating a missing docket or RIN as absent.
Inputs: no named parameters. Returns: object
get_rulemaking_docketRegulations.gov DocketARead-onlyInspect
One Regulations.gov docket, rulemaking or non-rulemaking: agency, type, title, abstract, the RIN when the agency entered one, when it was last modified, how many documents we hold and of which types, and the comment count when we measured it (null means not measured, never zero). Current state only; no comment text. Takes a docket_id such as EPA-HQ-OAR-2021-0208. An unknown docket is a 404 and is refunded. Read the documents with list_docket_documents.
Inputs: docket_id (required), Idempotency-Key. Returns: object
get_rulemaking_rinRulemaking by RINARead-onlyInspect
Everything we hold for one Regulation Identifier Number (NNNN-AANN, such as 2060-AV16): the latest Unified Agenda entry, OIRA reviews, Regulations.gov dockets, Federal Register documents by role (proposed, final, correction), CFR parts, public laws and related RINs, each link with its basis and quality. An ambiguous link carries no pick. coverage lists what we do not hold. An unknown RIN is a 404 and is refunded. For edition history use list_rin_agenda.
Inputs: rin (required), Idempotency-Key. Returns: object
get_section_changesSection Change HistoryARead-onlyInspect
What our refreshes have observed changing on one section over time: when it was added, amended or removed, newest first. This is our capture history, not the publisher's -- an empty list means we recorded no change, never that the section was never amended. For the publisher's own history, read amendmentHistory on the section.
Inputs: act_id (required), limit, order, sinceId, +3 more. Returns: object
get_section_cited_bySections Citing This OneARead-onlyInspect
The USC and CFR sections whose text cross-references a given section: the inverse of the crossReferences already returned on a section lookup. Use to find where a definition or requirement is actually invoked, or to gauge how load-bearing a provision is across the code.
Inputs: act_id (required), limit, offset, Idempotency-Key. Returns: object
get_section_cross_stateSame Rule in Other StatesARead-onlyInspect
Provisions in OTHER states that address the same subject as a given state statute section, ranked by similarity. State statutes only. Use for fifty-state surveys, multi-jurisdiction compliance, or to check whether a client's home-state rule is typical or an outlier.
Inputs: act_id (required), limit, Idempotency-Key. Returns: object
get_section_definitionsDefined Terms in SectionARead-onlyInspect
The term definitions that govern a section, parsed from its chapter's definitions section. Use whenever a provision turns on a term of art ('covered entity', 'security', 'employer') and you need the statute's own definition rather than the ordinary meaning.
Inputs: act_id (required), Idempotency-Key. Returns: object
get_section_enactmentsSession Laws Affecting a SectionARead-onlyInspect
The state session laws that a publisher's own 'sections affected' tables list against one code section: which act added, amended or repealed it, the act's effective dates, and a sessionLawId to read with get_session_law. Never a complete amendment history: only acts and sessions we hold. Only some states are supported; any other section answers supported: false and is not charged. Empty with supported: true means no held table lists it.
Inputs: act_id (required), Idempotency-Key. Returns: object
get_section_neighborsAdjacent SectionsARead-onlyInspect
The sections immediately before and after a given section within its own chapter or code, in statutory order. Use to read a provision in context, to find a definitions or penalties sibling, or to check whether the operative language continues into the next section.
Inputs: act_id (required), limit, Idempotency-Key. Returns: object
get_section_rulemakingFederal Register Documents for a CFR SectionARead-onlyInspect
The Federal Register documents behind one CFR section, from the section's printed source note and from eCFR amendment dates. matchBasis says which signal matched and quality is verified only when both agree; a candidate is a lead, not a finding. noteIsStale means eCFR records an amendment newer than the note prints. Read coverage before treating a missing document as absent. A document with one valid RIN carries rulemakingUrl, a pointer to get_rulemaking_rin. CFR sections only.
Inputs: act_id (required), Idempotency-Key. Returns: object
get_sections_batchStatute Sections (Batch)ARead-onlyInspect
Metadata for up to 50 sections in one call, by a list of act_ids. Same fields as get_us_statute_section. Use instead of looping that tool when you already hold several act_ids, for example every result of one search. Sections that are not found are skipped and not charged.
Inputs: actIds (required), includeBody, Idempotency-Key. Returns: object
get_session_lawState Session Law RecordARead-onlyInspect
One US state session law's record: citation, session, approval and effective dates, code sections it affects, official sources and textStatus. v1 covers state acts; federal session laws are on the statute tools. Accepts a sessionLawId or a citation (adds the resolve price, kept even if it does not resolve). A withheld law is still returned, textStatus withheld with withheldReason: we hold its text but a measured defect keeps it from being served.
Inputs: session_law_id (required), jurisdiction, Idempotency-Key. Returns: object
get_session_law_bodyState Session Law TextARead-onlyInspect
The text of one US state session law exactly as printed, about 30,000 characters per page; follow nextCursor while hasMore is true, each page charged. Check amendatoryMarkup: if lost, struck language may read as law. A withheld or not-held law answers available: false with a reason and is not charged: withheld means we hold the text but a measured defect keeps it back, so read the source instead.
Inputs: session_law_id (required), cursor, version, startChar, +3 more. Returns: object
get_session_law_changesSession Law Change FeedARead-onlyInspect
A sync feed of the US state session-law registry: every law that began to be served, gained text, was corrected, unserved, withdrawn or merged, oldest first. Start with sinceId=0, then pass the returned cursor while hasMore is true. detectedAt is when we recorded the change, never an effective date. State laws only. An empty page is still charged. Read a law from the feed with get_session_law or get_session_laws_batch.
Inputs: limit, sinceId, jurisdiction, Idempotency-Key. Returns: object
get_session_laws_batchState Session Laws (Batch)ARead-onlyInspect
Read up to 50 US state session laws in one call, by sessionLawId or citation: the record get_session_law returns, for each. Every entry is answered: served laws in laws, each carrying the input it answers, and the rest in notFound with a status. Duplicates collapse first. includeBody adds the first page of each law's text at the body price. To identify citations use resolve_statute_citations_batch; use this to read the records.
Inputs: ids (required), includeBody, jurisdiction, Idempotency-Key. Returns: object
get_session_law_versionsState Session Law VersionsARead-onlyInspect
Every version of one STATE session law that we served or adopted, newest first, and which is current. A law's text can be corrected after it is first served, so the version read yesterday may not be the one a plain body call returns today. Lists when each was recorded, its length, hash and how the text was taken; carries no text. Read one version with get_session_law_body and its version. Takes an id, not a citation. Charged once per call; a law with a single version still answers.
Inputs: session_law_id (required), Idempotency-Key. Returns: object
get_us_statute_sectionStatute Section MetadataARead-onlyInspect
Metadata for one US statute, regulation or rule section by act_id: citation, title hierarchy, breadcrumb, amendment history, and links to HTML, PDF and XML. Does NOT include the section text -- use get_us_statute_section_text for that. Good for confirming you have the right section before paying for its full body.
Inputs: act_id (required), full, Idempotency-Key. Returns: object
get_us_statute_section_textStatute Section TextARead-onlyInspect
The full text of a US statute, regulation or rule section by act_id. Returns styled HTML (with cross-references and paragraph numbering as officially published) and plain text. Use when you need the actual statutory language to quote, draft against, or analyze rather than just cite. A section the publisher prints as images answers bodyUnavailable: true with image_only_source and no text, and is not charged.
Inputs: act_id (required), asOf, format, layout, +3 more. Returns: object
list_docket_documentsDocuments in a DocketARead-onlyInspect
The documents of one Regulations.gov docket, oldest first: the proposal, notices, final rule and supporting material, with comment periods and Federal Register document numbers. Metadata only, no document text and no comments. Filter by documentType, openForComment and a postedFrom/postedTo window. Charged per page; an unknown docket is a 404 and is refunded.
Inputs: docket_id (required), limit, cursor, postedTo, +4 more. Returns: object
list_rin_agendaUnified Agenda History of a RINARead-onlyInspect
The Unified Agenda history of one RIN: one entry per edition that carried it (Fall 1995 onward), newest first, each with the stage, priority and timetable the agency printed in that edition. Narrow with from and to (edition ids YYYYMM, such as 202510). Charged per page; a RIN no edition carries is a 404 and is refunded. Use get_rulemaking_rin for the linked dockets and Federal Register documents.
Inputs: rin (required), to, from, limit, +2 more. Returns: object
list_rulemaking_committee_recordsRules Committee RecordsARead-onlyInspect
List Judicial Conference Rules Committee records (minutes, agenda books, reports, rules packages, comments), newest first. Filter by committee, docType, year, ruleSet, rule (needs ruleSet) and docketId. Drafting history, not law. Public comments and suggestions (textWithheld: true) withhold title, filename, sourceUrl, itemUrl and text: they name private people. Charged per page; an empty page is refunded. Read one with get_rulemaking_committee_record.
Inputs: rule, year, limit, cursor, +5 more. Returns: object
list_rulemaking_docketsRegulations.gov DocketsARead-onlyInspect
List and filter Regulations.gov dockets, most recently modified first, by agencyId, docketType (Rulemaking or Nonrulemaking), rin, hasRin and a modifiedSince/modifiedUntil window. Charged per page; an empty page is still charged. nextCursor is bound to the filters that produced it, so pass it back with the same filters. Read one with get_rulemaking_docket.
Inputs: rin, limit, cursor, hasRin, +5 more. Returns: object
list_session_lawsState Session LawsARead-onlyInspect
List one US state's session laws (acts as enacted, v1 covers state acts only). jurisdiction is REQUIRED, a two-letter state code. Filter by session, approval date or effective date, for example everything taking effect on a date. Rows include laws whose text is not served (textStatus not_held or withheld); pass hasText=true for readable ones. Page with cursor = nextCursor. An empty page is still charged.
Inputs: jurisdiction (required), q, isLaw, limit, +14 more. Returns: object
list_statute_divisionsBrowse Code StructureARead-onlyInspect
Walk the statutory hierarchy: list the level directly below what you pass, in statutory order. Which level that is depends on the code, so read level and each node's kind and pass its identifier back as the filter that kind names, until isLeaf is true and actId is set. One listing can mix kinds. name is null when the publisher prints no heading. A filter that does not fit the corpus is a 422 before any charge; an empty level is refunded; nextCursor means page on.
Inputs: corpusType (required), act, code, part, +18 more. Returns: object
list_statutes_coverageCorpus CoverageARead-onlyInspect
Coverage matrix: every corpusType we hold and its per-jurisdiction section counts, split by recorded status in statusBreakdown. Use before answering a jurisdiction question. A corpusType missing from corpora means no documents of that type are held for it yet; counts show what is held, not completeness. Pass state (a two-letter code, or federal) to read one jurisdiction; the full matrix is about a megabyte.
Inputs: state. Returns: object
resolve_statute_citationResolve CitationARead-onlyInspect
Resolve a Bluebook citation string ('42 U.S.C. 1983', '16 C.F.R. 444.1', 'Cal. Code Regs. tit. 22, 76227') to the exact section, confirmed, with an official source link and the act_id. Use this whenever the user gives you a citation rather than a question -- it is far more reliable than searching for the citation text. matchLevel means section is NOT the exact unit cited. A shared acronym (G.S. 52-577) is resolved: false until you pass state.
Inputs: cite (required), state, corpusType, Idempotency-Key. Returns: object
resolve_statute_citations_batchResolve Citations (Batch)ARead-onlyInspect
Resolve up to 50 Bluebook citations in one call, returning the same confirmed section, official source link and act_id as the single-citation tool. Use when a document or answer cites several provisions: one call instead of a loop. Priced per citation, the same as one at a time, so it buys latency, not a discount.
Inputs: citations (required), state, corpusType, Idempotency-Key. Returns: object
searchSearch US LawARead-onlyInspect
Generic corpus search over US primary law, returning {id, title, url} records for citation. Present so this server works in clients that require the standard search/fetch pair. If you can call search_us_statutes, prefer it: it filters by jurisdiction, corpus, date and status, which this cannot. Pair with fetch to read a result.
Inputs: query (required). Returns: object
search_session_lawsSearch State Session LawsARead-onlyInspect
Keyword search (BM25, not semantic) over the ENACTED text of US state session laws, best first, each with the passage that matched and a snippet. Send plain words, a quoted phrase (verbatim), a citation, a code section or a bill number. Words must match the act's own vocabulary: teeth cleaning will not find dental hygienist. Struck language is never searched. There is no 'no good match' flag, so read the snippet. An empty result is still charged.
Inputs: query (required), limit, offset, series, +7 more. Returns: object
search_us_statutesSearch US StatutesARead-onlyInspect
Search US federal and state statutes, regulations, constitutions and court rules. Default ranks by meaning; matchType boolean returns only sections satisfying a terms-and-connectors expression, pageable to the end. A result's act_id feeds the other statute tools; never hand-build one. For a citation, use resolve_statute_citation. citationOutsideFilters: the cited section exists outside your filters. 422: change the request, never resend it unchanged. 503: retry with backoff.
Inputs: query (required), code, part, limit, +22 more. Returns: object
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelogโฆ
TDQS
Scored across 37 tools
Most tools target a distinct resource+action, and descriptions explicitly disambiguate risky pairs (get_rulemaking_rin vs list_rin_agenda vs list_docket_documents; get_us_statute_section vs get_us_statute_section_text). However, generic `search`/`fetch` deliberately duplicate `search_us_statutes`/`get_us_statute_section_text`, and there are five single-vs-batch pairs plus three 'changes' tools whose boundaries must be read carefully rather than inferred.
Nearly everything follows a verb_noun snake_case pattern (list_rulemaking_dockets, get_session_law_body, resolve_statute_citation). Deviations are minor but real: the bare verbs `search` and `fetch`, and inconsistent entity vocabulary where the same object is called 'statute', 'us_statute' and 'section' across tools (search_us_statutes, get_us_statute_section, get_section_neighbors, get_sections_batch).
37 tools is heavy, though the server genuinely spans three large domains (statutes, state session laws, rulemaking/drafting history). Several tools are arguably redundant surface area โ generic search/fetch aliases, three batch variants, and two coverage/billing tools โ that push the count past what a well-scoped set would need.
Coverage is unusually deep for the legal-research domain: search, count, hierarchy walk, metadata, full text, neighbors, definitions, cross-references, cited-by, cross-state comparison, amendment history, citation resolution, plus session-law and Federal Register/Regulations.gov lifecycles. Remaining gaps are content-scoped rather than operation-shaped โ e.g. no keyword search over rulemaking/FR document text and no comment text retrieval โ which an agent can work around via docket and RIN tools.
Maintenance
Related MCP Connectors
CourtListener MCP โ Free Law Project's CourtListener API (v4).
MCP for CourtListener: US federal and state opinions, dockets, judges, plus eCFR regulations.
An MCP server that provides congressional transcripts
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceAn MCP server that provides comprehensive US legislation.34 npm38MIT
- FlicenseAqualityDmaintenanceMCP server for Legalize.dev, enabling AI to query and understand legal frameworks across countries via a structured legislation API.131-
- AlicenseAqualityBmaintenanceAn MCP server for accessing Turkish legislation (laws, regulations, decrees) via the Adalet Bakanligi API, providing search, full-text retrieval, and structured citations.562 PyPI2Apache 2.0
- AlicenseAqualityBmaintenanceMCP server for free U.S. case-law and court-docket search via CourtListener, providing tools for opinion search, docket lookup, citation verification, and more.1011 npmMIT