feat: add LangGraph example workflows - #184
Conversation
Signed-off-by: Zhongxuan Wang <daniewang@nvidia.com>
WalkthroughAdded shared configuration and two native LangGraph examples: a per-user calculator workflow with MCP tools and a structured email-phishing analyzer. Added MCP server support, CLI entry points, documentation, and offline tests. ChangesNative LangGraph examples
Estimated code review effort: 3 (Moderate) | ~25 minutes Sequence Diagram(s)sequenceDiagram
participant CalculatorCLI
participant PerUserReactAgent
participant MCPMathServer
participant NIMChatModel
CalculatorCLI->>PerUserReactAgent: Invoke with user ID and message
PerUserReactAgent->>MCPMathServer: Load configured calculator tools
PerUserReactAgent->>NIMChatModel: Execute ReAct graph
NIMChatModel->>MCPMathServer: Call calculator tool
PerUserReactAgent-->>CalculatorCLI: Return final message
sequenceDiagram
participant AnalyzerCLI
participant EmailPhishingGraph
participant NIMChatModel
participant PhishingAssessment
AnalyzerCLI->>EmailPhishingGraph: Submit email content
EmailPhishingGraph->>NIMChatModel: Analyze phishing indicators
NIMChatModel-->>EmailPhishingGraph: Return structured assessment
EmailPhishingGraph->>PhishingAssessment: Validate and serialize result
EmailPhishingGraph-->>AnalyzerCLI: Print JSON assessment
Possibly related PRs
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 8
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@examples/langgraph/calculator_mcp.py`:
- Around line 101-108: Make the exposed retry settings functional or remove
them: in examples/langgraph/calculator_mcp.py lines 101-108, update ainvoke to
read and apply retry_parsing_errors and max_retries around calculator
invocation; in examples/langgraph/email_phishing_analyzer.py lines 56-66, apply
the same settings around structured-model invocation. Ensure both workflows
either honor the configured retry behavior or remove the unsupported
configuration keys.
- Around line 81-99: The graph_for initialization path can create duplicate
graphs for the same user under concurrent calls. Add per-user synchronization
around the cache-miss initialization, recheck _sessions after acquiring the
lock, and cache only the single successfully created graph; clear or remove any
failed initialization state so later calls can retry. Add a concurrent
regression test covering simultaneous graph_for calls for one user.
In `@examples/langgraph/config.py`:
- Around line 6-7: Update the first product reference in
examples/langgraph/config.py lines 6-7 to use “NVIDIA NeMo Fabric,” and rename
the advertised server in examples/langgraph/mcp_math_server.py line 16 to
“NVIDIA NeMo Fabric calculator example.”
In `@examples/langgraph/README.md`:
- Line 79: Update the product references in the README passages around
“following additional work,” including the corresponding references later in the
document, to use “NeMo Fabric” after the full introduction at line 16. Replace
any standalone capitalized “Fabric” used as the product name, while preserving
the existing wording otherwise.
- Around line 46-47: Update the README text describing user ID reuse to state
that conversation resumption is process-local and works only while the same
PerUserReactAgent remains alive; clarify that each CLI invocation creates a new
agent, _sessions, and InMemorySaver, so a new process has no prior state. Do not
imply cross-process resumption unless durable checkpointing is implemented
first.
- Line 34: Quote the <your-api-key> placeholder in both NVIDIA_API_KEY export
commands in the README so the shell treats it as a literal value; update each
command to use the same quoted placeholder syntax.
In `@examples/README.md`:
- Line 40: Update the “LangGraph examples” heading in the examples README to use
title case: “LangGraph Examples.”
In `@tests/examples/test_langgraph_examples.py`:
- Around line 23-101: Extend the tests around load_config and PerUserReactAgent
to cover unknown llm_name, unknown tool source, blank user_id, and missing
configured MCP tools, asserting each raises the intended error. Add a concurrent
same-user graph_for test that verifies synchronization returns the same cached
graph and avoids duplicate model/client creation, and include lifecycle cleanup
assertions where the changed public API requires it.
🪄 Autofix
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Enterprise
Run ID: a7126ed0-ce40-411c-8e6a-5cae977431d9
📒 Files selected for processing (10)
examples/README.mdexamples/langgraph/README.mdexamples/langgraph/__init__.pyexamples/langgraph/calculator_mcp.pyexamples/langgraph/config.pyexamples/langgraph/configs/calculator_mcp.yamlexamples/langgraph/configs/email_phishing_analyzer.yamlexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/mcp_math_server.pytests/examples/test_langgraph_examples.py
📜 Review details
🧰 Additional context used
📓 Path-based instructions (25)
**/*.{rs,py,pyi,json,yaml,yml}
📄 CodeRabbit inference engine (.agents/skills/contribute-api/SKILL.md)
Determine and update every affected public surface, including the CLI, PyO3 bindings, Python SDK, type stubs, schemas, and adapter contract, so they remain in parity.
Files:
examples/langgraph/__init__.pyexamples/langgraph/configs/calculator_mcp.yamltests/examples/test_langgraph_examples.pyexamples/langgraph/mcp_math_server.pyexamples/langgraph/configs/email_phishing_analyzer.yamlexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
**/*
📄 CodeRabbit inference engine (.agents/skills/karpathy-guidelines/SKILL.md)
**/*: Before implementing, explicitly state assumptions, surface ambiguity and tradeoffs, present multiple interpretations when relevant, and ask for clarification rather than silently deciding or proceeding when requirements are unclear.
Prefer the minimum code needed to solve the requested problem: avoid speculative features, unnecessary abstractions, unrequested flexibility, and handling of impossible scenarios; simplify overcomplicated solutions.
When editing existing code, make surgical changes only: do not modify unrelated code, comments, formatting, or pre-existing dead code; match the existing style, and remove only unused imports, variables, or functions introduced by your changes.
Define verifiable success criteria for each task, such as writing regression tests for bugs and invalid-input tests for validation, then verify the implementation against those criteria. For multi-step work, state a brief plan with a verification check for each step.
**/*: Always spellNVIDIAin all caps; do not useNvidia,nvidia,nVidia,nVIDIA, orNV.
Usean NVIDIAbefore a noun, because the name begins with an “en” sound.
Do not add a registered trademark symbol afterNVIDIAwhen referring to the company; use trademark symbols with product names only when required by the document type or legal guidance.
Verify official capitalization, spacing, hyphenation, and spelling for NVIDIA and third-party product names; do not rewrite official product names for grammar or title-case rules.
Precede NVIDIA product names withNVIDIAon first mention when natural and accurate, and link the first mention when the destination helps the reader.
On first use, include the company name and full model qualifier when it helps identify the model; preserve official capitalization and punctuation, and use shorter family names only after establishing the full name.
For learning-oriented and developer content, do not force trademark symbols unless explicitly required; for press, ...
Files:
examples/langgraph/__init__.pyexamples/langgraph/configs/calculator_mcp.yamltests/examples/test_langgraph_examples.pyexamples/langgraph/mcp_math_server.pyexamples/README.mdexamples/langgraph/configs/email_phishing_analyzer.yamlexamples/langgraph/README.mdexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
**/*.{rs,py}
📄 CodeRabbit inference engine (.agents/skills/prepare-pr/SKILL.md)
For native binding changes, run
cargo check -p fabric-python --locked.Use
snake_casefor functions and variables; usePascalCasefor Rust types and Python classes.
Files:
examples/langgraph/__init__.pytests/examples/test_langgraph_examples.pyexamples/langgraph/mcp_math_server.pyexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
**/*.{py,pyi}
📄 CodeRabbit inference engine (.agents/skills/validate-change/SKILL.md)
If Python code or a Python-facing adapter changes, run
just test-python.In Python SDK, adapters, examples, and tests, follow the existing style, use type annotations for public APIs, and keep native binding declarations synchronized with their Rust implementations.
Files:
examples/langgraph/__init__.pytests/examples/test_langgraph_examples.pyexamples/langgraph/mcp_math_server.pyexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
**/*.{rs,py,pyi}
📄 CodeRabbit inference engine (.agents/skills/validate-change/SKILL.md)
**/*.{rs,py,pyi}: If public configuration types change, confirm schema snapshot tests injust test-rustpass and review generated schema diffs.
For schema or public contract changes, run both language suites and review changes underschemas/and generated API references.
Files:
examples/langgraph/__init__.pytests/examples/test_langgraph_examples.pyexamples/langgraph/mcp_math_server.pyexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
**/*.{py,pyi,rs}
📄 CodeRabbit inference engine (.agents/skills/validate-change/SKILL.md)
For Python SDK or PyO3 binding changes, use
python-tests, run focused pytest tests first, thenjust test-python; rebuild withjust build-pythonwhen native code or packaging changes.Public contract changes must keep native Python binding declarations synchronized with their Rust implementations.
Files:
examples/langgraph/__init__.pytests/examples/test_langgraph_examples.pyexamples/langgraph/mcp_math_server.pyexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
**/*.{rs,py,toml}
📄 CodeRabbit inference engine (.agents/skills/update-project-version/SKILL.md)
When editing version helpers, verify every
nemo-fabric-*workspace package through Cargo metadata and reject a static version inpython/pyproject.toml.
Files:
examples/langgraph/__init__.pytests/examples/test_langgraph_examples.pyexamples/langgraph/mcp_math_server.pyexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
**/*.{toml,rs,py}
📄 CodeRabbit inference engine (.agents/skills/update-project-version/SKILL.md)
Avoid blind repository-wide replacement of version-like strings; distinguish package-version references from examples and unrelated dependency versions.
Files:
examples/langgraph/__init__.pytests/examples/test_langgraph_examples.pyexamples/langgraph/mcp_math_server.pyexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
examples/{README.md,**/*}
📄 CodeRabbit inference engine (AGENTS.md)
Update examples documentation and relevant examples when public behavior, the package, or supported bindings change.
Files:
examples/langgraph/__init__.pyexamples/langgraph/configs/calculator_mcp.yamlexamples/langgraph/mcp_math_server.pyexamples/README.mdexamples/langgraph/configs/email_phishing_analyzer.yamlexamples/langgraph/README.mdexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
**/*.{md,mdx,yml,py,rs,sh}
📄 CodeRabbit inference engine (.agents/skills/review-doc-style/SKILL.md)
Keep documentation aligned with current NeMo Fabric behavior, repository layout, entry points, commands, package names, APIs, bindings, and support claims.
Files:
examples/langgraph/__init__.pytests/examples/test_langgraph_examples.pyexamples/langgraph/mcp_math_server.pyexamples/README.mdexamples/langgraph/README.mdexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
{adapters/**,examples/**}
⚙️ CodeRabbit configuration file
{adapters/**,examples/**}: Review adapter and example changes for command correctness, config/schema consistency, artifact handling, and compatibility with the public NeMo Fabric contracts.
Files:
examples/langgraph/__init__.pyexamples/langgraph/configs/calculator_mcp.yamlexamples/langgraph/mcp_math_server.pyexamples/README.mdexamples/langgraph/configs/email_phishing_analyzer.yamlexamples/langgraph/README.mdexamples/langgraph/config.pyexamples/langgraph/email_phishing_analyzer.pyexamples/langgraph/calculator_mcp.py
**/*.{yml,yaml,toml,lock}
📄 CodeRabbit inference engine (.agents/skills/validate-change/SKILL.md)
For CI or packaging changes, use
maintain-ciormaintain-packaging, then run recipes and checks whose behavior changed.
Files:
examples/langgraph/configs/calculator_mcp.yamlexamples/langgraph/configs/email_phishing_analyzer.yaml
**/*.{toml,yaml,yml,sh,bash}
📄 CodeRabbit inference engine (CONTRIBUTING.md)
TOML, YAML, and shell files must use the specified
#SPDX copyright and Apache-2.0 license headers.
Files:
examples/langgraph/configs/calculator_mcp.yamlexamples/langgraph/configs/email_phishing_analyzer.yaml
tests/**/*.{rs,py}
📄 CodeRabbit inference engine (CONTRIBUTING.md)
When adding functionality, include tests in the corresponding Rust crate or the relevant area under
tests/.
Files:
tests/examples/test_langgraph_examples.py
tests/**/*.py
📄 CodeRabbit inference engine (.agents/skills/python-tests/SKILL.md)
tests/**/*.py: Use pytest to run Python tests.
Do not add@pytest.mark.asyncioto tests; async tests are automatically detected by the async runner.
Do not add-> Nonereturn type annotations to test functions.
When mocking a class, useunittest.mock.MagicMockorAsyncMock, using thespecargument when necessary, rather than defining a new class.
Prefix mocked class names withmock, notfake.
Prefer pytest fixtures over helper methods.
If a fixture is needed in multiple test files, define it once inconftest.pyrather than repeating it.
Define fixtures using@pytest.fixture(name="<fixture_name>"[, scope="<scope>"])and a<fixture_name>_fixturefunction; specifyscopeonly when it is notfunction.
Preferpytest.mark.parametrizeover separate tests for different input types.
Use@pytest.mark.usefixtureswhen a fixture is needed but its returned value is unused or it returns no value.
Avoid defensive programming in tests; access expected values directly so missing data raises a clear failure, such as usingresults["data"]instead ofresults.get("data").
When adapter installation metadata changes, packaging metadata tests must directly assert that the root project depends unconditionally on the exact-versionnemo-fabric-runtimedistribution.
Packaging metadata tests must verify that each root harness extra delegates to the matching version of the leaf adapter'sharnessextra.
Packaging metadata tests must verify that bare leaf dependencies remain adapter-owned and that the rootadapter-testsdependency group installs each leaf through itsharnessextra.
Packaging metadata tests must verify that every leaf providesfull; only adapters importing NeMo Relay Python APIs providerelay, while adapters using an external Relay executable havefullequal toharness.
Files:
tests/examples/test_langgraph_examples.py
{tests/**,python/tests/**}
⚙️ CodeRabbit configuration file
{tests/**,python/tests/**}: Tests should cover the behavior promised by the changed API surface, including error paths, lifecycle cleanup, and SDK/native parity where relevant.
Files:
tests/examples/test_langgraph_examples.py
**/*.{md,rst}
📄 CodeRabbit inference engine (.agents/skills/contribute-api/SKILL.md)
Update documentation and examples in the same branch as the public API change.
Files:
examples/README.mdexamples/langgraph/README.md
{README.md,docs/**/*.{md,mdx,yml},examples/**/*.{md,mdx,yml}}
📄 CodeRabbit inference engine (.agents/skills/contribute-docs/SKILL.md)
Keep package names, repository references, and build commands current in documentation and examples.
Files:
examples/README.mdexamples/langgraph/README.md
{docs/**/*.{md,mdx,yml},examples/**/*.{md,mdx,yml}}
📄 CodeRabbit inference engine (.agents/skills/contribute-docs/SKILL.md)
Update relevant getting-started, reference, adapter, and example documentation when the corresponding examples or adapters change.
Files:
examples/README.mdexamples/langgraph/README.md
**/*.{md,mdx,rst}
📄 CodeRabbit inference engine (.agents/skills/review-doc-style/assets/nvidia-style-guide.md)
**/*.{md,mdx,rst}: For NeMo Fabric documentation, verify technical claims against the current repository, public API, or documented command before reviewing style.
Always spellNVIDIAin all caps; do not useNvidia,nvidia, orNV.
Format commands, code elements, expressions, package names, file names, and paths as inline code.
Use descriptive link text; avoid raw URLs and weak anchors such ashereorread more.
Use title case consistently for technical documentation headings.
Introduce code blocks, lists, tables, and images with complete sentences.
Write procedures as imperative, parallel steps; split long procedures into smaller tasks.
Prefer active voice, present tense, short sentences, contractions, and plain English while preserving necessary technical precision.
Usecanfor possibility and reservemayfor permission.
Useafterfor temporal relationships instead ofonce, and preferrefer tooverseewhen directing readers to another resource.
Avoid culture-specific idioms, unnecessary Latinisms, jokes, and marketing exaggeration in technical documentation.
Spell out months in body text, avoid ordinal dates, and use clear time zones.
Spell out whole numbers from zero through nine unless they are technical values, parameters, versions, or UI values; use numerals for 10 or greater and commas in thousands.
Do not add trademark symbols to learning-oriented documentation unless the source, platform, or legal guidance explicitly requires them.
Do not replace precise technical terms with simpler words when doing so would lose precision.
Do not flag passive voice when the actor is unknown or the action is the important part.
Do not rewrite API names, package names, command flags, or code literals for style.
**/*.{md,mdx,rst}: Use consistent title case for technical-document headings and table headers; avoid quotation marks, ampersands, and exclamation marks in headings, while preserving official product, event, research, and whitepaper title ...
Files:
examples/README.mdexamples/langgraph/README.md
**/*.{md,rst,txt,adoc}
📄 CodeRabbit inference engine (.agents/skills/review-doc-style/assets/nvidia-style-language-mechanics.md)
**/*.{md,rst,txt,adoc}: For technical documentation, use professional, active, conversational, engaging, precise, and plain-English prose. Prefer active voice, present tense, short sentences, and scannable paragraphs. Avoid casual or imprecise language, swearing, threats, insults, jokes, puns, culture-specific idioms, marketing exaggeration, and unsupported third-party comparisons.
Usecanfor possibility and reservemayfor permission; useafterfor temporal order; userefer tofor cross-references; prefer short direct sentences and specific verbs; avoid unnecessarypleasein technical documentation.
Prefer active voice when the actor matters. Passive voice is acceptable when the actor is unknown or irrelevant, when the action or result is the focus, or in programmer documentation.
Use natural contractions in conversational technical prose, but do not force them in formal legal copy, API references, or generated text.
Prefer simpler English over Latinisms: usefor exampleorsuch asinstead ofe.g.,and so oninstead ofetc.,that isinstead ofi.e.,compared toinstead ofvs., andby,through, orusinginstead ofvia. Use industry-standard terms such as in silico, in vitro, and in vivo when appropriate, and italicize them in running text.
Usethatwithout commas for essential clauses, andwhichwith commas for nonessential clauses.
Format dates and times clearly: spell out months in body text; use forms such asJune 12, 2025; avoid numeric or ordinal dates; capitalize days; use 12-hour time when appropriate; include a space beforea.m.orp.m.; useETandPTfor needed time zones; avoid24/7; and preferfrom 12:30 to 1:00 p.m.for prose ranges.
Format numbers consistently: spell out zero through nine in body text, use numerals for 10 or greater and for technical values, use commas in thousands, do not begin a sentence with a numeral, spell out ordinals, and use numerals consistently within a category wh...
Files:
examples/README.mdexamples/langgraph/README.md
**/*.{md,mdx}
📄 CodeRabbit inference engine (.agents/skills/review-doc-style/SKILL.md)
**/*.{md,mdx}: Use the full product nameNVIDIA NeMo Fabricon first use, typically in the title and H1; useNeMo Fabricthereafter. Usefabricalone only for the CLI tool and surround it with backticks.
Treat incorrect or stale commands, package names, paths, APIs, support claims, procedures, examples, terminology, or public behavior documentation as blocking issues.
CapitalizeNVIDIAcorrectly and format code, commands, paths, and filenames as inline code where needed.
Use title case for technical-documentation headings.
Introduce code blocks, tables, and lists with complete lead-in sentences; ensure examples match current APIs and build commands.
Use descriptive anchor text, avoid raw URLs and generic labels such ashere, and use repository-relative.mdxpaths for links withindocs/.
Prefer active voice, present tense, short sentences, plain English, consistent terminology, and imperative, parallel, scannable procedures.
Useafterinstead ofoncewhen expressing temporal sequence, and usecanrather thanmaywhen describing possibility rather than permission.
Avoid ambiguous numeric dates and ordinal dates in body text.
For learning-oriented documentation, do not force trademark symbols unless the source document explicitly requires them.
When reporting documentation-review findings, lead withMust fix,Should fix, andNice to havecategories; include file path, line reference, current problem, rationale, and a concrete rewrite or direction.
Files:
examples/README.mdexamples/langgraph/README.md
**/*.md
📄 CodeRabbit inference engine (CONTRIBUTING.md)
Update relevant SDK, API reference, adapter, example, integration, and support documentation when the corresponding public surface changes.
Files:
examples/README.mdexamples/langgraph/README.md
**/*.{html,md}
📄 CodeRabbit inference engine (CONTRIBUTING.md)
HTML and Markdown files must use the specified SPDX HTML-comment header.
Files:
examples/README.mdexamples/langgraph/README.md
{*.md,**/*.md,**/*.mdx,**/*.ipynb}
⚙️ CodeRabbit configuration file
{*.md,**/*.md,**/*.mdx,**/*.ipynb}: Enforce the product name in user-facing prose: use "NVIDIA NeMo Fabric" on first use and "NeMo Fabric" thereafter. Flag standalone capitalized "Fabric" when it refers to the product. Do not flag the lowercasefabricCLI command, package/import/crate names, code identifiers, API symbols, configuration keys, file paths, or unrelated generic uses of the word.
Files:
examples/README.mdexamples/langgraph/README.md
🪛 ast-grep (0.45.0)
examples/langgraph/email_phishing_analyzer.py
[info] 87-87: use jsonify instead of json.dumps for JSON output
Context: json.dumps(result["assessment"], indent=2)
Note: [CWE-116] Improper Encoding or Escaping of Output.
(use-jsonify)
🪛 Ruff (0.16.1)
examples/langgraph/mcp_math_server.py
[warning] 41-41: Avoid specifying long messages outside the exception class
(TRY003)
examples/langgraph/config.py
[warning] 77-77: Remove quotes from type annotation
Remove quotes
(UP037)
[warning] 80-80: Avoid specifying long messages outside the exception class
(TRY003)
[warning] 86-86: Avoid specifying long messages outside the exception class
(TRY003)
[warning] 110-110: Prefer TypeError exception for invalid type
(TRY004)
[warning] 110-110: Avoid specifying long messages outside the exception class
(TRY003)
[warning] 119-121: Avoid specifying long messages outside the exception class
(TRY003)
examples/langgraph/email_phishing_analyzer.py
[warning] 50-50: Dynamically typed expressions (typing.Any) are disallowed in model
(ANN401)
[warning] 51-51: Dynamically typed expressions (typing.Any) are disallowed in build_email_phishing_analyzer
(ANN401)
[warning] 55-55: Avoid specifying long messages outside the exception class
(TRY003)
examples/langgraph/calculator_mcp.py
[warning] 64-64: Avoid specifying long messages outside the exception class
(TRY003)
[warning] 66-66: Avoid specifying long messages outside the exception class
(TRY003)
[warning] 68-68: Avoid specifying long messages outside the exception class
(TRY003)
[warning] 76-76: Dynamically typed expressions (typing.Any) are disallowed in graph_for
(ANN401)
[warning] 80-80: Avoid specifying long messages outside the exception class
(TRY003)
🔇 Additional comments (4)
examples/langgraph/config.py (1)
17-22: 🩺 Stability & AvailabilityNo dependency declaration issue.
The documented setup installs these dependencies through the
adapters,adapter-tests, andtestgroups.> Likely an incorrect or invalid review comment.examples/langgraph/__init__.py (1)
1-4: LGTM!examples/langgraph/README.md (1)
1-18: LGTM!Also applies to: 20-33, 35-45, 48-56, 58-78, 80-87, 89-94, 96-108
examples/README.md (1)
42-45: LGTM!
| graph = self._sessions.get(user_id) | ||
| if graph is not None: | ||
| return graph | ||
|
|
||
| server = self._config.mcp.servers["mcp_math"] # validated in __init__ | ||
| client = self._mcp_client_factory( | ||
| {"mcp_math": mcp_connection(server)}, tool_name_prefix=False | ||
| ) | ||
| mcp_tools = _selected_mcp_tools(list(await client.get_tools()), server) | ||
| model = self._model_factory(self._config.selected_model()) | ||
| graph = self._graph_factory( | ||
| model, | ||
| [current_timezone, *mcp_tools], | ||
| checkpointer=InMemorySaver(), | ||
| debug=bool(self._config.workflow.settings.get("verbose", False)), | ||
| name="per_user_calculator", | ||
| ) | ||
| self._sessions[user_id] = graph | ||
| return graph |
There was a problem hiding this comment.
🩺 Stability & Availability | 🟠 Major | ⚡ Quick win
Synchronize first graph creation for each user.
Two concurrent graph_for(user_id) calls can both miss _sessions, await client.get_tools(), and create different graphs. The later call overwrites the cached graph. The earlier call then uses a separate checkpoint store for the same user.
Use a per-user initialization lock or shared initialization task. Recheck _sessions after acquiring the lock. Remove failed initialization state. Add a concurrent regression test.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@examples/langgraph/calculator_mcp.py` around lines 81 - 99, The graph_for
initialization path can create duplicate graphs for the same user under
concurrent calls. Add per-user synchronization around the cache-miss
initialization, recheck _sessions after acquiring the lock, and cache only the
single successfully created graph; clear or remove any failed initialization
state so later calls can retry. Add a concurrent regression test covering
simultaneous graph_for calls for one user.
| async def ainvoke(self, user_id: str, message: str) -> dict[str, Any]: | ||
| """Run a message in the graph and persisted conversation for ``user_id``.""" | ||
|
|
||
| graph = await self.graph_for(user_id) | ||
| return await graph.ainvoke( | ||
| {"messages": [{"role": "user", "content": message}]}, | ||
| {"configurable": {"thread_id": user_id}}, | ||
| ) |
There was a problem hiding this comment.
🎯 Functional Correctness | 🟠 Major | 🏗️ Heavy lift
Make retry settings functional or remove them.
Both YAML configurations expose retry_parsing_errors and max_retries, but neither workflow consumes them. Users cannot obtain the configured retry behavior.
examples/langgraph/calculator_mcp.py#L101-L108: Read and apply the retry settings around calculator invocation, or remove the unsupported settings.examples/langgraph/email_phishing_analyzer.py#L56-L66: Read and apply the retry settings around structured-model invocation, or remove the unsupported settings.
📍 Affects 2 files
examples/langgraph/calculator_mcp.py#L101-L108(this comment)examples/langgraph/email_phishing_analyzer.py#L56-L66
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@examples/langgraph/calculator_mcp.py` around lines 101 - 108, Make the
exposed retry settings functional or remove them: in
examples/langgraph/calculator_mcp.py lines 101-108, update ainvoke to read and
apply retry_parsing_errors and max_retries around calculator invocation; in
examples/langgraph/email_phishing_analyzer.py lines 56-66, apply the same
settings around structured-model invocation. Ensure both workflows either honor
the configured retry behavior or remove the unsupported configuration keys.
| This is application configuration for the examples, not a NeMo Fabric adapter | ||
| descriptor or a supported Fabric configuration format. |
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win
Use NVIDIA NeMo Fabric on first reference.
examples/langgraph/config.py#L6-L7: Change the first product reference toNVIDIA NeMo Fabric.examples/langgraph/mcp_math_server.py#L16-L16: Change the advertised server name toNVIDIA NeMo Fabric calculator example.
As per coding guidelines, “Precede NVIDIA product names with NVIDIA on first mention when natural and accurate.”
📍 Affects 2 files
examples/langgraph/config.py#L6-L7(this comment)examples/langgraph/mcp_math_server.py#L16-L16
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@examples/langgraph/config.py` around lines 6 - 7, Update the first product
reference in examples/langgraph/config.py lines 6-7 to use “NVIDIA NeMo Fabric,”
and rename the advertised server in examples/langgraph/mcp_math_server.py line
16 to “NVIDIA NeMo Fabric calculator example.”
Source: Coding guidelines
| Set an NVIDIA API key and start the included MCP server in one terminal: | ||
|
|
||
| ```bash | ||
| export NVIDIA_API_KEY=<your-api-key> |
There was a problem hiding this comment.
🎯 Functional Correctness | 🔴 Critical | ⚡ Quick win
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
set -u
file="examples/langgraph/README.md"
printf '%s\n' '--- relevant lines ---'
sed -n '29,60p' "$file"
printf '%s\n' '--- shell syntax checks ---'
for command in \
'export NVIDIA_API_KEY=<your-api-key>' \
'export NVIDIA_API_KEY="<your-api-key>"'
do
if bash -n <(printf '%s\n' "$command") 2>/tmp/bash-syntax.err; then
printf 'PASS: %s\n' "$command"
else
printf 'FAIL: %s\n' "$command"
cat /tmp/bash-syntax.err
fi
doneRepository: NVIDIA/NeMo-Fabric
Length of output: 1445
🏁 Script executed:
#!/bin/bash
set -u
for command in \
'export NVIDIA_API_KEY=<your-api-key>' \
'export NVIDIA_API_KEY="<your-api-key>"'
do
err=$(mktemp)
if printf '%s\n' "$command" | bash -n 2>"$err"; then
printf 'PASS: %s\n' "$command"
else
printf 'FAIL: %s\n' "$command"
cat "$err"
fi
rm -f "$err"
doneRepository: NVIDIA/NeMo-Fabric
Length of output: 356
Quote the API key placeholders in both shell commands.
export NVIDIA_API_KEY=<your-api-key> is invalid shell syntax because < is parsed as redirection. Use export NVIDIA_API_KEY="<your-api-key>" at lines 34 and 57.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@examples/langgraph/README.md` at line 34, Quote the <your-api-key>
placeholder in both NVIDIA_API_KEY export commands in the README so the shell
treats it as a literal value; update each command to use the same quoted
placeholder syntax.
Source: Coding guidelines
| Each user ID creates a separate graph, `InMemorySaver`, and | ||
| `MultiServerMCPClient`. Reusing a user ID resumes only that user's conversation. |
There was a problem hiding this comment.
🗄️ Data Integrity & Integration | 🟠 Major | ⚡ Quick win
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
set -euo pipefail
ast-grep outline examples/langgraph/calculator_mcp.py --items all
rg -n -C 8 'PerUserReactAgent|InMemorySaver|asyncio\.run|def main|__main__|user-id' \
examples/langgraph/calculator_mcp.pyRepository: NVIDIA/NeMo-Fabric
Length of output: 4476
🏁 Script executed:
#!/bin/bash
set -euo pipefail
printf '%s\n' '--- README context ---'
cat -n examples/langgraph/README.md | sed -n '1,75p'
printf '%s\n' '--- lifecycle and state references ---'
rg -n -C 6 '_sessions|InMemorySaver|thread_id|build_per_user_react_agent|asyncio\.run|--user-id|ainvoke' \
examples/langgraph tests 2>/dev/null || true
printf '%s\n' '--- source section ---'
cat -n examples/langgraph/calculator_mcp.py | sed -n '52,145p'Repository: NVIDIA/NeMo-Fabric
Length of output: 47813
Must fix: Document process-local conversation resumption.
State that reusing --user-id resumes a conversation only while the same PerUserReactAgent remains alive. Each CLI invocation creates a new agent, _sessions, and InMemorySaver, so a new process starts without prior conversation state. Add durable checkpointing before documenting cross-process resumption.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@examples/langgraph/README.md` around lines 46 - 47, Update the README text
describing user ID reuse to state that conversation resumption is process-local
and works only while the same PerUserReactAgent remains alive; clarify that each
CLI invocation creates a new agent, _sessions, and InMemorySaver, so a new
process has no prior state. Do not imply cross-process resumption unless durable
checkpointing is implemented first.
Source: Coding guidelines
| ## Work Needed for a Full Adapter | ||
|
|
||
| The examples validate the two workflow shapes, but a generic adapter needs the | ||
| following additional work before it can support them through NVIDIA NeMo Fabric: |
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win
Should fix: Use NeMo Fabric after the first full mention.
In examples/langgraph/README.md, Lines 79, 88, and 95 repeat NVIDIA NeMo Fabric or use standalone Fabric for the product. Line 16 already introduces the full name.
Proposed wording
- before it can support them through NVIDIA NeMo Fabric:
+ before it can support them through NeMo Fabric:
- Map a selected Fabric model alias to an NIM binding.
+ Map a selected NeMo Fabric model alias to an NIM binding.
- The calculator shows per-user state, but Fabric currently scopes a runtime
+ The calculator shows per-user state, but NeMo Fabric currently scopes a runtimeAs per coding guidelines, use the full product name on first use and NeMo Fabric thereafter. As per path instructions, do not use standalone capitalized Fabric for the product.
Also applies to: 88-88, 95-95
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@examples/langgraph/README.md` at line 79, Update the product references in
the README passages around “following additional work,” including the
corresponding references later in the document, to use “NeMo Fabric” after the
full introduction at line 16. Replace any standalone capitalized “Fabric” used
as the product name, while preserving the existing wording otherwise.
Sources: Coding guidelines, Path instructions
| --input "Reply with exactly: NeMo Fabric works" | ||
| ``` | ||
|
|
||
| ## LangGraph examples |
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win
Should fix: Use title case for the new heading.
In examples/README.md, Line 40 uses ## LangGraph examples. Change it to ## LangGraph Examples.
As per coding guidelines, technical-documentation headings must use title case.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@examples/README.md` at line 40, Update the “LangGraph examples” heading in
the examples README to use title case: “LangGraph Examples.”
Source: Coding guidelines
| def test_calculator_config_preserves_requested_mcp_and_workflow_shape(): | ||
| config = load_config(CALCULATOR_CONFIG) | ||
|
|
||
| assert config.selected_model().model == "meta/llama-3.1-70b-instruct" | ||
| assert config.mcp is not None | ||
| assert config.mcp.servers["mcp_math"].transport == "streamable-http" | ||
| assert config.workflow.entrypoint == "langgraph:per_user_react_agent" | ||
| assert config.workflow.settings["tool_names"] == ["current_timezone", "mcp_math"] | ||
|
|
||
|
|
||
| async def test_calculator_creates_isolated_graph_and_mcp_client_per_user(): | ||
| config = load_config(CALCULATOR_CONFIG) | ||
| mock_tools = [] | ||
| for name in config.mcp.servers["mcp_math"].include: # validated by the example | ||
| mock_tool = MagicMock() | ||
| mock_tool.name = name | ||
| mock_tools.append(mock_tool) | ||
| mock_client_factory = MagicMock() | ||
| mock_client_factory.side_effect = [ | ||
| MagicMock(get_tools=AsyncMock(return_value=mock_tools)) | ||
| for _ in range(2) | ||
| ] | ||
| mock_graph_factory = MagicMock(side_effect=[MagicMock(), MagicMock()]) | ||
| mock_model_factory = MagicMock(side_effect=[MagicMock(), MagicMock()]) | ||
|
|
||
| agent = PerUserReactAgent( | ||
| config, | ||
| model_factory=mock_model_factory, | ||
| mcp_client_factory=mock_client_factory, | ||
| graph_factory=mock_graph_factory, | ||
| ) | ||
| alice_first = await agent.graph_for("alice") | ||
| alice_second = await agent.graph_for("alice") | ||
| hatter = await agent.graph_for("hatter") | ||
|
|
||
| assert alice_first is alice_second | ||
| assert alice_first is not hatter | ||
| assert mock_model_factory.call_count == 2 | ||
| assert mock_client_factory.call_count == 2 | ||
| connection = mock_client_factory.call_args.args[0]["mcp_math"] | ||
| assert connection == { | ||
| "transport": "streamable_http", | ||
| "url": "http://localhost:9901/mcp", | ||
| } | ||
| for call in mock_graph_factory.call_args_list: | ||
| assert call.kwargs["checkpointer"] is not None | ||
| assert call.kwargs["name"] == "per_user_calculator" | ||
|
|
||
|
|
||
| def test_current_timezone_uses_the_explicit_server_configuration(restore_environ): | ||
| restore_environ["TZ"] = "America/Los_Angeles" | ||
|
|
||
| assert current_timezone.invoke({}) == "America/Los_Angeles" | ||
|
|
||
|
|
||
| async def test_phishing_graph_projects_a_json_safe_structured_assessment(): | ||
| config = load_config(PHISHING_CONFIG) | ||
| mock_structured_model = MagicMock() | ||
| mock_structured_model.ainvoke = AsyncMock( | ||
| return_value=PhishingAssessment( | ||
| is_likely_phishing=True, | ||
| explanation="It asks for banking information to complete a refund.", | ||
| ) | ||
| ) | ||
| mock_model = MagicMock() | ||
| mock_model.with_structured_output.return_value = mock_structured_model | ||
|
|
||
| graph = build_email_phishing_analyzer(config, model=mock_model) | ||
| result = await graph.ainvoke( | ||
| {"body": "Provide your routing number so we can issue a refund."} | ||
| ) | ||
|
|
||
| assert result["assessment"] == { | ||
| "is_likely_phishing": True, | ||
| "explanation": "It asks for banking information to complete a refund.", | ||
| } | ||
| mock_model.with_structured_output.assert_called_once_with( | ||
| PhishingAssessment, method="function_calling" | ||
| ) |
There was a problem hiding this comment.
🎯 Functional Correctness | 🟡 Minor | ⚡ Quick win
Add error-path coverage for the new public configuration and workflows.
Add tests for an unknown llm_name, unknown tool source, blank user_id, and missing configured MCP tools. Add a concurrent same-user test for graph_for after synchronization is added.
As per path instructions, tests should cover changed API behavior, including error paths and lifecycle cleanup where relevant.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@tests/examples/test_langgraph_examples.py` around lines 23 - 101, Extend the
tests around load_config and PerUserReactAgent to cover unknown llm_name,
unknown tool source, blank user_id, and missing configured MCP tools, asserting
each raises the intended error. Add a concurrent same-user graph_for test that
verifies synchronization returns the same cached graph and avoids duplicate
model/client creation, and include lifecycle cleanup assertions where the
changed public API requires it.
Source: Path instructions
1 participant
Overview
Adds two runnable native LangGraph examples: a per-user calculator with an MCP math server and a structured email-phishing analyzer. These examples run directly with LangGraph. NVIDIA NeMo Fabric does not yet ship a LangGraph adapter, and the example YAML files are not FabricConfig input.
Details
Validation
Where should the reviewer start?
Start with examples/langgraph/README.md for the support boundary and adapter requirements, then review the two graph factories and their focused offline tests.
Related Issues: (use one of the action keywords Closes / Fixes / Resolves / Relates to)
Relates to FABRIC-166
I confirm this contribution is my own work, or I have the right to submit it under this project's license.
I searched existing issues and open pull requests, and this does not duplicate existing work.
Summary by CodeRabbit