> ## Documentation Index
> Fetch the complete documentation index at: https://docs.esheria.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Esheria

> Contextualised, computable legal knowledge for product, compliance, and agent workflows.

Esheria gives applications contextualised, computable legal knowledge: discover current packs, compare versions, monitor legal changes, retrieve obligations, test applicability, build filing calendars and evidence registers, inspect penalties and review audit metadata, query relationships, and verify claims with citation spans.

Every response preserves the audit trail that regulatory products need: `trace_id`, readiness labels, limitation disclosures, citation IDs, quote spans, and publication mode.

<Info>
  Esheria returns source-backed regulatory workflow data, not legal advice. Keep legal review in the loop for client-facing advice.
</Info>

## Legal Operating Layer

<Columns cols={3}>
  <Card title="Section-depth law" icon="landmark">
    Work with structured statutes, regulations, obligations, filing rules, evidence requirements, penalties, review metadata, and citation-backed source context.
  </Card>

  <Card title="Workflow-ready APIs" icon="workflow">
    Turn legal data into applicability reviews, registers, calendars, evidence checklists, change monitors, and product surfaces without losing the source trail.
  </Card>

  <Card title="Agent-safe tools" icon="bot">
    Give assistants MCP tools that return citations, limitations, and `trace_id` with every legal workflow.
  </Card>
</Columns>

## Start Building

<Columns cols={2}>
  <Card title="Create a self-serve workspace" icon="user-plus" href="/guides/dashboard-self-serve">
    Sign up at `dashboard.esheria.ai`, create a data token, manage credits, and test calls in the playground.
  </Card>

  <Card title="Start in five minutes" icon="rocket" href="/quickstart">
    Set a dashboard-created token, run the first `curl`, install `esheria`, configure MCP, and verify a claim with citations.
  </Card>

  <Card title="Open the API reference" icon="braces" href="/api-reference/operations/healthz">
    Use generated OpenAPI pages and the playground for every public endpoint.
  </Card>

  <Card title="Monitor legal changes" icon="radar" href="/guides/change-monitoring">
    Compare pack versions, inspect diffs, and route change events into downstream review.
  </Card>

  <Card title="Run regulatory intelligence" icon="network" href="/guides/regulatory-intelligence">
    Monitor sources, rebuild graph coverage, run customer applicability, and track customer impacts.
  </Card>

  <Card title="Map evidence and filings" icon="clipboard-check" href="/guides/filing-evidence-and-penalties">
    Use first-class filing rules, evidence requirements, penalty facts, and legal review audit metadata.
  </Card>

  <Card title="Use the command line" icon="terminal" href="/cli">
    Run smoke checks, list packs, export evidence, and inspect claim-verification output.
  </Card>

  <Card title="Connect agent tools" icon="bot" href="/agent-tools/mcp">
    Expose regulatory tools to Claude Desktop, Cursor, Codex MCP, or an internal host.
  </Card>

  <Card title="Explore legal use cases" icon="briefcase-business" href="/solutions/use-cases">
    Map Esheria workflows to law firm, fintech, corporate registry, GRC, HR, tax, and AI product scenarios.
  </Card>
</Columns>

## First Request

```bash theme={null}
export ESHERIA_API_BASE_URL="https://api.esheria.ai"
export ESHERIA_API_KEY="YOUR_DASHBOARD_CREATED_DATA_TOKEN"

curl --compressed -sS \
  "$ESHERIA_API_BASE_URL/api/v1/domain-packs?readiness_label=verified_published&limit=10" \
  -H "x-api-key: $ESHERIA_API_KEY" | jq
```

The response shape is stable across endpoints:

```json theme={null}
{
  "status": "ok",
  "data": {
    "packs": []
  },
  "errors": [],
  "trace_id": "request-trace-id"
}
```

Errors use the same envelope:

```json theme={null}
{
  "status": "error",
  "data": null,
  "errors": [
    {
      "code": "unauthorized",
      "message": "Invalid API key"
    }
  ],
  "trace_id": "request-trace-id"
}
```

## Operational Model

<Columns cols={3}>
  <Card title="Published facts only" icon="shield-check">
    Customer-facing endpoints serve the published regulatory pack tier and disclose readiness limitations.
  </Card>

  <Card title="Citation-first output" icon="quote">
    Obligations, evidence, calendars, relationships, and claims carry citation IDs and quote-span context.
  </Card>

  <Card title="Traceable results" icon="scan-search">
    API and CLI preserve the API envelope. MCP returns bounded structured content with `trace_id` and truncation metadata.
  </Card>
</Columns>

## Core Workflows

<Columns cols={2}>
  <Card title="Discover published packs" icon="package-search" href="/guides/packs">
    API, CLI, and MCP entry points for current regulatory packs.
  </Card>

  <Card title="Retrieve obligations" icon="list-checks" href="/guides/obligations">
    Published duties with classification, actionability, actors, workflow targets, citations, and quote spans.
  </Card>

  <Card title="Check applicability" icon="route" href="/recipes/examples#applicability-check">
    Match obligations to structured entity profile facts, activities, sectors, and processing context.
  </Card>

  <Card title="Monitor changes" icon="radar" href="/guides/change-monitoring">
    Use versions, diffs, and change events before updating tenant operating records.
  </Card>

  <Card title="Build filing and evidence workflows" icon="clipboard-check" href="/guides/filing-evidence-and-penalties">
    Filing rules, evidence requirements, penalties, and review audit metadata with source trace.
  </Card>

  <Card title="Verify generated claims" icon="badge-check" href="/guides/claim-verification">
    Test model or user claims against published facts before display.
  </Card>

  <Card title="Connect agent hosts" icon="bot" href="/agent-tools/mcp">
    MCP tools for Claude Desktop, Cursor, Codex MCP, and internal hosts.
  </Card>
</Columns>

## First User Journey

<Steps>
  <Step title="Sign up or sign in">
    Open `https://dashboard.esheria.ai/`, sign up or sign in, and resolve your workspace.
  </Step>

  <Step title="Create a data token">
    Open **API Tokens**, create a data token, store the one-time secret, then set it as `ESHERIA_API_KEY`.
  </Step>

  <Step title="Run the first curl">
    Call `/healthz`, then list packs with `/api/v1/domain-packs`.
  </Step>

  <Step title="Install the CLI">
    Run `pip install esheria`, then `esheria packs list --format json`.
  </Step>

  <Step title="Configure MCP">
    Add the live `https://mcp.esheria.ai/mcp` URL, then use OAuth where supported or a dashboard data token as the bearer-token alternative.
  </Step>

  <Step title="Verify a claim">
    Use `verify-claim` to test a generated answer against published facts, citations, and quote spans.
  </Step>
</Steps>

## Current Pack Posture

The sanitized discovery release contains 69 pack records across 27 jurisdiction labels. 60 are `verified_published`; 9 remain metadata-only `not_ready` records. Catalog version: `2026-07-15-v1+sha256:c5732893a6156d077166b60ea5016166ac5b178bb024162320d3a54e0ce745eb`. Coverage includes Africa, Europe, North and South America, Asia-Pacific, Middle East, and global regulatory sources across privacy, financial crime, cybersecurity, tax, sustainability, virtual assets, AI, health, and other domains.

<Card title="View the production pack catalog" icon="package-search" href="/guides/pack-catalog">
  See every current pack ID, jurisdiction, legal domain, readiness state, and
  published-fact count.
</Card>

All listed packs expose the generic pack workflows for obligations,
applicability, filing calendars, evidence registers, penalty facts, legal
review audit, versions, diffs, change events, relationships, and export.
Claim verification is available only for packs with an enabled claim evaluation
profile; other packs return `unsupported_capability` for that workflow while
the generic pack endpoints continue to work.