# MaskFlow PII proxy [Health: Active] **Category:** ๐Ÿ’ป Developer Tools **Repository:** https://github.com/maskflow/maskflow **GitHub Stars:** 0 **Views:** 0 **Installs:** 0 **Upvotes:** 0 **Directory Page:** https://allmcps.com/mcp/maskflow-pii-proxy ## Description Mask PII in outbound MCP tool-call arguments, unmask the results. Indian identifiers included. ## Claude Desktop Quick Installation Install path detected from listing signals. Uses `npx` (confidence: high): ```json "mcpServers": { "maskflow-pii-proxy": { "command": "npx", "args": ["-y","@modelcontextprotocol/server-github"] } } ``` ## Documentation & README # MaskFlow **Stop Indian PII from ever reaching an LLM.** Aadhaar, PAN, GSTIN, UPI, IFSC, ABHA, Indian names and addresses โ€” detected and replaced with reversible, typed placeholders before a prompt leaves your process, restored in the response. 28 entity types, checksum-validated where a public checksum exists, MIT-licensed, runs entirely on your own infrastructure. [![CI](https://github.com/maskflow/maskflow/actions/workflows/ci.yml/badge.svg)](https://github.com/maskflow/maskflow/actions/workflows/ci.yml) [![PyPI](https://img.shields.io/pypi/v/maskflow-sdk)](https://pypi.org/project/maskflow-sdk/) [![npm](https://img.shields.io/npm/v/%40maskflow%2Fdetection)](https://www.npmjs.com/package/@maskflow/detection) [![License: MIT](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE) [![Python 3.10+](https://img.shields.io/badge/python-3.10%2B-blue)](pyproject.toml)

Terminal demo: pip install maskflow-sdk, then mask() replaces an Aadhaar number and email with <AADHAAR_1> and <EMAIL_1> before an LLM call, and unmask() restores the originals in the response

## Why India's DPDP Act sets a compliance deadline of **13 May 2027**, with penalties of up to **โ‚น250 crore** for a breach where the required safeguards weren't in place. Every prompt sent to an LLM provider is a potential data-sharing event โ€” and general-purpose PII tools weren't built to recognize Aadhaar, PAN, GSTIN, UPI VPAs, IFSC codes, ABHA health IDs, or Indian names and addresses reliably. [Presidio](https://github.com/microsoft/presidio) already owns generic PII and is more mature everywhere else; MaskFlow exists specifically to close that gap, with accuracy that's measured and published, not asserted. See [MaskFlow vs. alternatives](#maskflow-vs-alternatives). ## Quickstart ```bash pip install maskflow-sdk python -m spacy download en_core_web_sm ``` ```python from maskflow import mask, unmask result = mask("My Aadhaar is 2346 8907 6543 and you can reach me at alice@example.com.") result.masked_text # "My Aadhaar is and you can reach me at ." unmask(result.masked_text, result.mapping) # original text, restored ``` For a one-line wrapper around your actual LLM call, or session-scoped masking across a multi-turn agent (same value โ†’ same token for as long as the session is open), see [`packages/maskflow-sdk/README.md`](https://github.com/maskflow/maskflow/blob/HEAD/packages/maskflow-sdk/README.md). ## How it works 1. **Tier-0 excision first.** Deterministic regex/checksum matches (Aadhaar, PAN, GSTIN, email, credit card, ...) are found and locked in *before* the NER pass ever runs โ€” spaCy parses each document at most once, only over what tier-0 didn't already claim. 2. **Every match is a `Span`.** Start/end offsets, entity type, confidence, which recognizer produced it, whether a checksum validated it, and a human-readable explanation trail. Run `maskflow explain ""` (from `maskflow-cli`) to see that trail for any input, span by span โ€” including near-misses that fell just below threshold and what config change would catch them. 3. **Deterministic resolution on overlaps.** Below-threshold spans are dropped; among what's left, a checksum-validated span always beats an overlapping unvalidated one, then higher confidence, then longer span, then earliest start wins โ€” greedy, non-overlapping placement. 4. **Placeholders are typed, stable, and collision-proof.** ``, ``, ... โ€” the same value gets the same token within a session, and if the input text already contains something that looks like a placeholder, a nonce suffix (``) is used instead so a real placeholder is never ambiguous with attacker-controlled input. 5. **Recognizers are pluggable.** `maskflow-pack-intl` and `maskflow-pack-india` are just two `"maskflow.recognizers"` entry-point plugins sharing one memoised analysis context โ€” write and register your own the same way. See [`docs/custom-recognizers.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/custom-recognizers.md). ## Protecting your own logs Regex/checksum-based recognizers can also scrub your application's own `logging` calls โ€” not just text passed through `mask()` โ€” closing the gap where a raw value gets logged before it's ever masked: ```python from maskflow_core import install_pii_filter install_pii_filter() # attaches to the root logger, once, at startup ``` Opt-in only; importing `maskflow_core` never touches global logging state on its own. It doesn't cover NER-only entity types (bare names/addresses) or `exc_info` tracebacks โ€” see [`docs/logging.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/logging.md) for the exact boundary. ## Auditing what already reached a provider Going forward, `mask()` keeps PII out of your prompts. But the DPDP audit asks a backward-looking question first: *what has this system already sent to a third-party LLM?* `maskflow scan` answers it. It reads your historical LLM traffic โ€” a JSONL/CSV export, a recursive directory, an S3 archive, a Postgres table, or the Langfuse / Helicone / LangSmith API โ€” streams it through the same detection with bounded memory (parallel, resumable), and writes **one self-contained HTML report**: a single headline number, breakdowns by entity type / provider / model / time, a severity ranking with a plain-English "why this matters" per row, **masked excerpts only** (never a raw value), and a DPDP Rule 6 mapping appendix. Also `--format json|csv`. Runs entirely locally โ€” the API sources only *read* from your own account, nothing is transmitted. ```bash pipx install maskflow-cli # or: docker run --rm -v "$PWD:/work" ghcr.io/maskflow/cli maskflow scan jsonl requests.jsonl --field 'messages[].content' --deep -o exposure.html ``` Also ships as a standalone binary (mac/linux/windows, no Python โ€” pattern pass only) and a [GitHub Action](https://github.com/maskflow/maskflow/blob/HEAD/packaging/scan-action/) that can fail a build over a PII-exposure threshold. A runnable 60-record synthetic example is in [`packages/maskflow-cli/examples/`](https://github.com/maskflow/maskflow/blob/HEAD/packages/maskflow-cli/examples/); full reference, including the Rule 6 mapping, in [`docs/scan.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/scan.md). ## Gateway: no code change at all `maskflow-gateway` is a drop-in OpenAI/Anthropic-compatible proxy. Point your existing client's base URL at it and PII is masked before every request reaches the provider and restored in the response โ€” **streaming included** (a `` split across SSE chunks is stitched back together; fuzz-tested at every byte boundary). Tool-call arguments are walked as JSON; multi-turn token identity is kept in Redis (AES-256-GCM at rest). ```python from openai import OpenAI client = OpenAI(base_url="http://localhost:8000/v1", api_key="sk-...") # your real key, passed through ``` ```bash pip install "maskflow-gateway[redis]" # or: docker run -p 8000:8000 ghcr.io/maskflow/gateway ``` Full reference in [`packages/maskflow-gateway/README.md`](https://github.com/maskflow/maskflow/blob/HEAD/packages/maskflow-gateway/README.md) and [`docs/gateway.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/gateway.md). ## LiteLLM: a guardrail on your existing proxy Already running a [LiteLLM](https://github.com/BerriAI/litellm) proxy? `maskflow-litellm` is a custom guardrail โ€” no separate service. It masks PII (Indian identifiers included) before a request leaves the proxy and restores it in the response, streaming and tool calls included. ```bash pip install maskflow-litellm ``` ```yaml guardrails: - guardrail_name: maskflow litellm_params: guardrail: maskflow_litellm.MaskflowGuardrail mode: [pre_call, post_call] ``` Full reference in [`packages/maskflow-litellm/README.md`](https://github.com/maskflow/maskflow/blob/HEAD/packages/maskflow-litellm/README.md) and [`docs/litellm-guardrail.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/litellm-guardrail.md). ## LangChain: a one-line import swap `maskflow-langchain` is a drop-in for [`langchain-experimental`](https://github.com/langchain-ai/langchain-experimental)'s Presidio anonymizer โ€” same `.anonymize()` / `.deanonymize()` / `.deanonymizer_mapping` โ€” so an existing chain migrates by changing one import. The deanonymizer is a streaming-aware `Runnable` (a placeholder split across streamed chunks is stitched back), and there's an optional leak-guard callback that fails a call closed if PII reaches the model. ```bash pip install maskflow-langchain ``` ```python # from langchain_experimental.data_anonymizer import PresidioReversibleAnonymizer from maskflow_langchain import MaskflowReversibleAnonymizer as PresidioReversibleAnonymizer ``` Full reference in [`packages/maskflow-langchain/README.md`](https://github.com/maskflow/maskflow/blob/HEAD/packages/maskflow-langchain/README.md) and [`docs/langchain.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/langchain.md). ## LlamaIndex: keep PII out of RAG `maskflow-llamaindex` gives a LlamaIndex RAG pipeline two components and an unmask helper. `MaskflowNodePostprocessor` is a drop-in for `llama_index.core.postprocessor.PIINodePostprocessor` (same `__pii_node_info__` contract) that masks retrieved context before the synthesizer, with no LLM call. `MaskflowIngestionTransform` masks node text at ingestion so raw PII never reaches the vector store. `unmask_response()` / `MaskflowQueryEngine` restore the real values in the answer. ```bash pip install maskflow-llamaindex ``` ```python from maskflow_llamaindex import MaskflowNodePostprocessor, unmask_response engine = index.as_query_engine(node_postprocessors=[MaskflowNodePostprocessor()]) response = engine.query("What is Ramesh's PAN?") answer = unmask_response(str(response), response.source_nodes) ``` Full reference in [`packages/maskflow-llamaindex/README.md`](https://github.com/maskflow/maskflow/blob/HEAD/packages/maskflow-llamaindex/README.md) and [`docs/llamaindex.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/llamaindex.md). ## MCP: a masking proxy for agent tool calls `maskflow-mcp` is a [Model Context Protocol](https://modelcontextprotocol.io) proxy. Put it in front of any MCP server and PII in outbound `tools/call` arguments is masked before it reaches the tool, results are unmasked on the way back, and placeholders stay consistent for the whole agent run. Agent tooling is where PII leakage is least examined; this is a drop-in shim that stops the real values at the boundary. ```jsonc { "mcpServers": { "github": { "command": "maskflow-mcp", "args": ["stdio", "--backend", "npx -y @modelcontextprotocol/server-github", "--pass-env", "GITHUB_TOKEN"] } } } ``` Full reference in [`packages/maskflow-mcp/README.md`](https://github.com/maskflow/maskflow/blob/HEAD/packages/maskflow-mcp/README.md) and [`docs/mcp.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/mcp.md). ### Evidence โ€” a record of what was masked `maskflow-evidence` emits a **metadata-only** record of *what* was masked โ€” entity type, count, recognizer, action, versions โ€” and never a value, a placeholder, or the mapping. Off by default; one line in `.maskflowrc` turns it on, to a self-hosted sink (stdout / file / syslog / webhook / OTLP). The gateway emits automatically when enabled; `maskflow explain --evidence` does it for a single run. ```toml [evidence] enabled = true sink = "file" path = "evidence.log" ``` That no event field can carry free text is enforced in CI. Full reference in [`docs/evidence.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/evidence.md). Compliance-control mapping and signed accuracy attestations (R5 items 2โ€“3) are still being validated with practitioners and are not yet part of this layer. ## Configuration Drop a `.maskflowrc` (TOML/YAML/JSON) in your project to adjust entity thresholds, disable an entity, add a custom regex-based entity, exclude specific values, or change the substitution strategy per entity (`replace` / `redact` / `mask` / `hash` / `surrogate` โ€” the last swaps in a plausible *fake* value drawn from reserved/test-only ranges, e.g. RFC 2606 example domains or publicly documented payment-industry test card numbers, instead of a placeholder token). `mask()`/`mask_and_call()`/`session()` all pick it up automatically; no `.maskflowrc` anywhere behaves exactly as before this existed: ```toml [entities.PHONE] strategy = "mask" # "415-555-0132" -> "XXX-XXX-0132" instead of "" [custom.EMPLOYEE_ID] pattern = '\bEMP-\d{6}\b' score = 0.9 ``` ```bash pip install maskflow-cli # maskflow config validate / maskflow config show --resolved ``` See [`docs/configuration.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/configuration.md) for the full schema and precedence rules. ## What it detects today `maskflow-sdk` and `maskflow-cli` both bundle `maskflow-pack-intl` and `maskflow-pack-india` โ€” installing either gets you all 28 types below with no extra install step. **International (12 types)** โ€” `maskflow-pack-intl`: | Type | How | |---|---| | Email | Regex | | Phone | Regex | | SSN | Regex + area-code validation | | Credit card | Regex + Luhn checksum | | IP address (v4/v6) | Regex | | AWS access key | Regex | | API key / generic secret | Regex | | JWT | Regex | | IBAN | Regex + mod-97 checksum | | Street address | Regex | | Person name | spaCy NER | | Date of birth | spaCy NER + keyword context | **Indian (17 types, the moat)** โ€” `maskflow-pack-india`: | Type | How | |---|---| | Aadhaar (UID + VID) | Regex + Verhoeff checksum | | Aadhaar (masked display form, e.g. `XXXX XXXX 9012`) | Regex, unvalidated, needs context | | PAN | Regex + holder-category structural check (no public final-letter checksum) | | GSTIN | Regex + state-code range + embedded-PAN check + base-36 checksum | | IFSC | Regex + bank code against a bundled RBI code list | | UPI VPA | Regex + PSP handle against a bundled NPCI handle list | | Indian mobile number | Regex, full confidence with a `+91`/`0` prefix, needs context otherwise | | PIN code | Regex, unvalidated, needs context (pin/pincode/state name/address) | | Voter ID (EPIC number) | Regex, structural only (no public checksum) | | Indian passport number | Regex, structural only (no public checksum) | | Indian passport MRZ block | Regex + 4 ICAO 9303 check digits | | Driving licence | Regex + state RTO code against a bundled code list | | Vehicle registration | Regex + state RTO code against a bundled code list | | ABHA number (health ID) | Regex, unvalidated, needs context | | ABHA address | Regex + domain (abdm/sbx) check | | Bank account number (India) | Regex, unvalidated, needs context (account/a/c/acct) | | Person name (Indian) | Gazetteer (name corpus) + structural (honorifics, relational markers, initials, form fields) + spaCy NER agreement boost | | Indian address | Gazetteer (554+ Indian cities/places) + structural (unit markers, landmark-relative phrasing, locality patterns) | (`PERSON_NAME` is one shared entity type produced by both packs' layers, so 12 + 17 โˆ’ 1 shared = 28 unique types total.) ## Multi-language `@maskflow/detection` on npm is a TypeScript port of the 10 pure regex/structural intl types (email, phone, SSN, credit card, IP, AWS key, API key, JWT, IBAN, street address) โ€” same API shape as the Python SDK, tested against the same fixtures so both stay accuracy-matched. `PERSON_NAME`/`DATE_OF_BIRTH` (need spaCy NER) and the India pack's checksum-validated types are Python-only โ€” `@maskflow/detection` is a deliberately narrow browser/Node helper, not a second full engine. See [`packages/maskflow-js/README.md`](https://github.com/maskflow/maskflow/blob/HEAD/packages/maskflow-js/README.md). ```ts import { mask, unmask } from "@maskflow/detection"; const result = mask("Email me at alice@example.com or call 415-555-0132."); unmask(result.maskedText, result.mapping); // original text, restored ``` ## MaskFlow vs. alternatives | | MaskFlow | Presidio | mask-privacy | |---|---|---|---| | Indian identifiers with checksums | Aadhaar, PAN, GSTIN, IFSC, UPI (in `maskflow-sdk`) | No | No | | Session-consistent tokens (unmask later) | Yes | Via custom anonymizer config | Yes, today | | NER | spaCy | spaCy, Stanza, transformers | Regex-based, no NER | | Breadth / maturity | Narrow, early (28 types) | Broad, mature (Microsoft-backed, years of production use) | Narrow, early | | License | MIT | MIT | Varies by package | | Languages | Python + TypeScript (regex layer) | Python, multi-language via configurable NLP models | JS/TS | Presidio is ahead on breadth and maturity. If you need broad, battle-tested coverage today, use it. MaskFlow's bet is Indian-identifier accuracy and a reversible mask/unmask flow that's simpler to drop into a single, provider-agnostic call. ## Benchmark Real numbers, not vendor claims โ€” including results where competitors beat us. Scored on [`indiapii-v1.0`](https://github.com/maskflow/maskflow/blob/HEAD/bench/indiapii/data/indiapii-v1.0.jsonl), 2000 synthetic, checksum-valid documents (Aadhaar/PAN/GSTIN pass the same validity math the real formats use), against stock Presidio, Presidio with two hand-added Aadhaar/PAN recognizers, and mask-privacy. F1 below is partial-overlap matching (exact-character matching is too strict for multi-token spans like addresses โ€” see the full report for both). | Entity | MaskFlow | Presidio (stock) | Presidio + custom | mask-privacy | |---|---|---|---|---| | GSTIN / IFSC / UPI VPA | 100% | not supported | not supported | not supported | | AADHAAR | 98.4% | not supported | 96.6% | not supported | | PAN | 100% | not supported | 100% | not supported | | Indian mobile number | 99.0% | 94.9% | 94.9% | 41.9% | | Person name | 47.3% | 30.4% | 30.4% | 37.7% | | Indian address | 43.3% | 48.2% | 48.2% | **57.9%** | Indian address is the one row above where a competitor is ahead โ€” our gazetteer still has room to grow, and we're not hiding that. Full per-entity breakdown (all 17 types), strict-vs-partial matching, and latency/memory numbers: [`bench/reports/indiapii-v1.0/results.md`](https://github.com/maskflow/maskflow/blob/HEAD/bench/reports/indiapii-v1.0/results.md). Reproduce with `uv sync --group bench && uv run python -m bench.indiapii.harness run`; harness source in [`bench/indiapii/harness/`](https://github.com/maskflow/maskflow/blob/HEAD/bench/indiapii/harness/). ### International / US-shaped types Generic PII is not a surface MaskFlow is built to win โ€” Presidio and mask-privacy own it โ€” but "measured, not asserted" applies to the intl pack too. Scored on [`intl-pii-v1.0`](https://github.com/maskflow/maskflow/blob/HEAD/bench/intlpii/data/intl-pii-v1.0.jsonl), 1800 synthetic documents (Luhn-valid cards, mod-97-valid IBANs, SSNs in real-but-unassigned area ranges), same harness, partial-overlap F1: | Entity | MaskFlow | Presidio (stock) | mask-privacy | |---|---|---|---| | Email / IP address | 100% | 100% | 100% | | Phone | 100% | 86.7% | 64.1% | | Credit card | 99.4% | **100%** | 78.2% | | IBAN | 99.4% | 76.3% | **100%** | | AWS key / API key / JWT | 100% | not supported | not supported | | SSN | 100% | 100% | not detected at defaults | | Street address | 99.8% | 6.5%ยน | 81.1% | | Person name | 74.0% | **84.1%** | **86.1%** | | Date of birth | 63.6% | 29.5%ยน | **79.8%** | Competitors are ahead on **person name** (MaskFlow's NER recognizer over-trusts spaCy's `PERSON` tag on sentence-initial words) and, for mask-privacy, on **date of birth** (it ships a dedicated birth-date regex; MaskFlow's spaCy-only `DATE` pass misses numeric formats and the pack's context-keyword list omits "birth date"). Both are tracked follow-ups. ยน `LOCATION` / `DATE_TIME` are mapped generously to `ADDRESS` / `DATE_OF_BIRTH` so those engines score non-zero at all โ€” see [`bench/intlpii/harness/labels.py`](https://github.com/maskflow/maskflow/blob/HEAD/bench/intlpii/harness/labels.py). Full table (all 12 types, strict + partial, latency/memory): [`bench/reports/intl-pii-v1.0/results.md`](https://github.com/maskflow/maskflow/blob/HEAD/bench/reports/intl-pii-v1.0/results.md). Reproduce with `uv sync --group bench && uv run python -m bench.intlpii.harness run`. ### Does masking hurt the LLM's answer? [`bench/indiapii/quality`](https://github.com/maskflow/maskflow/blob/HEAD/bench/indiapii/quality) is a 200-task, LLM-judged benchmark that runs each task unmasked, with placeholder masking, and with surrogate masking, then measures the masked-minus-unmasked delta in task completion, fluency, factual consistency, and field-extraction accuracy (plus a hard leak check โ€” a `` token must never survive `unmask()` into the final answer). Method and scoring are built and unit-tested; a **published run is pending** an `ANTHROPIC_API_KEY` (`make quality-bench` โ€” ~1200 disk-cached calls, ~$1.50). Once run, `bench/reports/indiapii-quality-v1.0/results.md`. ### `maskflow scan` on log-shaped input The detection benches above are prose; [`maskflow scan`](https://github.com/maskflow/maskflow/blob/HEAD/docs/scan.md) runs over access-log lines, JSON app logs, stack traces, and request dumps. [`scan-log-v1.0`](https://github.com/maskflow/maskflow/blob/HEAD/bench/scanbench/data/) is 2000 synthetic log records with the PII-lookalike noise real logs carry (trace ids, UUIDs, git SHAs, internal IPs, `File.java:142` frames), scoring detection and โ€” the metric that matters for an audit โ€” the **false-positive rate**. `--deep`, partial-overlap F1: | Entity | MaskFlow | Presidio (stock) | Naive regex | |---|---|---|---| | Aadhaar / PAN / GSTIN / IFSC / UPI | 100% | not supported | 67โ€“100% | | Email / Indian mobile | 100% | 79โ€“85% | 95โ€“100% | | Credit card | 97.8% | 100% | 86.2% | | Person name | 87.9% | 57.6% | not supported | | IP address | 66.7%ยน | 66.7%ยน | 59.6%ยน | **False positives / 1000 log records:** MaskFlow `--deep` 428, its patterns-only pass 298, Presidio 544, naive regex 433. Checksum-validated identifiers essentially never false-positive on log noise (naive regex flags 337 fake Aadhaars; MaskFlow flags 0); the `--deep` NER pass produces ~2.5ร— the false `PERSON_NAME` hits that the patterns pass does. ยน every internal `10./172.16./192.168.` IP is reported as `IP_ADDRESS` โ€” a known gap (no private-IP suppression). A plumbing check verifies the real scan pipeline surfaces exactly what `detect()` finds. Full report: [`bench/reports/scan-log-v1.0/results.md`](https://github.com/maskflow/maskflow/blob/HEAD/bench/reports/scan-log-v1.0/results.md). ## Roadmap Openly not done yet, so you know what you're signing up for: - `maskflow-gateway` hardening: more provider schemas, a Redis-cluster session backend, and first-class OpenTelemetry traces. ## Links - Site: [maskflow.in](https://maskflow.in) - Docs: [`docs/configuration.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/configuration.md), [`docs/custom-recognizers.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/custom-recognizers.md), [`docs/scan.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/scan.md), [`docs/dpdp-rule6.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/dpdp-rule6.md), [`docs/agent-sessions.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/agent-sessions.md), [`docs/logging.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/logging.md), [`docs/gateway.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/gateway.md), [`docs/litellm-guardrail.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/litellm-guardrail.md), [`docs/langchain.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/langchain.md), [`docs/llamaindex.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/llamaindex.md), [`docs/mcp.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/mcp.md), [`docs/data-refresh.md`](https://github.com/maskflow/maskflow/blob/HEAD/docs/data-refresh.md) - [Changelog](https://github.com/maskflow/maskflow/blob/HEAD/CHANGELOG.md) - [Security policy](https://github.com/maskflow/maskflow/blob/HEAD/SECURITY.md) - [Contributing](https://github.com/maskflow/maskflow/blob/HEAD/CONTRIBUTING.md) - License: [MIT](https://github.com/maskflow/maskflow/blob/HEAD/LICENSE)