Eco Policy MCP

eco-policy-mcp

An MCP server for Indian finance data — mutual-fund NAVs, GST & income-tax calculators, RBI policy rates and NSE market data — built for any MCP client (Claude, Cursor, Gemini CLI, VS Code, Cline, custom agents...), not just one.

Every successful response carries a provenance block (source, as_of, reference), so AI agents can cite where each number came from instead of hallucinating it.

Links

Resource	URL
GitHub repository	https://github.com/Akhilesh-Chandewar/eco-policy-mcp
PyPI package	https://pypi.org/project/eco-policy-mcp/
MCP Registry name	io.github.Akhilesh-Chandewar/eco-policy-mcp (active on registry.modelcontextprotocol.io)
Releases	https://github.com/Akhilesh-Chandewar/eco-policy-mcp/releases
Issues	https://github.com/Akhilesh-Chandewar/eco-policy-mcp/issues
License	MIT

Install & run

# Run with any MCP client via uvx (installs automatically from PyPI):
uvx eco-policy-mcp

# Or pin the version:
uvx eco-policy-mcp@0.1.0

# From a clone:
uv sync
uv run eco-policy-mcp            # stdio (default, local clients)
uv run eco-policy-mcp --http --port 8000   # streamable-HTTP (remote clients)

See examples/ for client configs (Claude Desktop, Cursor, Gemini CLI, VS Code, remote HTTP).

Tools

Domain	Tool	What it does
GST	gst_split	CGST/SGST vs IGST breakup for an amount + slab rate
GST	gst_from_inclusive	Extract GST from a tax-inclusive amount
GST	gst_reverse_charge	Reverse-charge (RCM) tax payable
GST	gstin_validate	GSTIN mod-36 checksum validation + state code
Income tax	income_tax_new_regime	FY 2026-27 new-regime estimate (std deduction + 87A rebate)
Income tax	income_tax_old_regime	Old-regime estimate (80C, age-based exemption)
Income tax	income_tax_compare_regimes	New vs old with recommendation
Mutual funds	mf_search_schemes	Search all AMFI schemes by name
Mutual funds	mf_get_nav	Latest NAV for a scheme code
Mutual funds	mf_get_nav_history	Historical NAVs between dates
Mutual funds	mf_compute_returns	Absolute return + CAGR between dates
Mutual funds	mf_sip_calculator	SIP simulation on real NAVs with XIRR
RBI	rbi_get_policy_rates	Repo, SDF, MSF, bank rate, CRR, SLR (as-of dated)
NSE	nse_get_quote	Equity quote (unofficial public endpoint)
NSE	nse_get_index_quote	Index snapshot (NIFTY 50, BANK, IT...)
NSE	nse_get_historical	Daily OHLCV (max 1y window)
Server	eco_server_info	Version, tool list, cache health

Also exposes resources (eco-policy://amfi/freshness, eco-policy://rbi/policy-rates) and prompts (compare_mutual_funds, gst_invoice_breakup).

Data sources & disclaimers

• AMFI (NAVAll.txt) — official daily NAV dump; parsed and cached 6h.
• mfapi.in — community mirror of AMFI's NAV archive (history, returns, SIP).
• RBI — curated policy-rate snapshot stamped with its as-of date; verify at rbi.org.in before acting on it.
• NSE — unofficial public endpoints; quotes may be delayed and endpoints can change without notice. Failures return structured errors, never raw HTML.
• Tax computations are estimators (excludes surcharge, capital-gains special rates); nothing here is investment or tax advice.

Configuration

All settings via env vars (prefix ECO_): ECO_TTL_AMFI_INDEX, ECO_TTL_NSE_QUOTE, ECO_NSE_MIN_INTERVAL, ECO_HTTP_TIMEOUT, ECO_MFAPI_BASE, ECO_NSE_BASE, ECO_AMFI_NAV_URL, ... — see src/eco_policy_mcp/config.py.

Development

uv sync --dev          # install with dev tools
uv run pytest          # run the test suite (no network needed)

License

MIT — see LICENSE.