Atlas / MCP servers / redpanda-data / Property Extractor Unit Tests

Property Extractor Unit TestsBLOCK

Name: Property Extractor Unit Tests
Author: redpanda-data

mcp/redpanda-data/docs-extensions-and-macros

MCP server by redpanda-data · Audited 2026-10-11

⚑ Report / re-scan Source ↗

Extensions and macros developed for Redpanda documentation.

Verdict

BLOCK

Grade

Trust score

52 /100

Exposed tools

131 125r · 4w · 2d

Transport

stdio

License

—

Stars

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

Overview

Comprehensive unit tests for the Redpanda property extraction system. These tests validate the transformer pipeline that processes C++ property declarations into JSON documentation format.

Test Coverage

EnterpriseTransformer Tests (5 tests)

test_restricted_only_with_enum_vector: Validates correct classification when a property has restricted values AND an enum definition vector (enum is superset)
test_restricted_with_sanctioned_disjoint_values: Validates classification when restricted and sanctioned values are disjoint sets
test_restricted_only_single_vector: Tests simple restricted_only pattern with a single restriction vector
test_non_enterprise_property: Ensures non-enterprise properties are correctly flagged
test_simple_enterprise_property: Tests simple enterprise properties with lambda validators

ParamNormalizerTransformer Tests (3 tests)

test_normalize_restriction_at_position_0: Validates normalization when restriction vector is at parameter position 0
test_normalize_restriction_at_position_1: Validates normalization when restriction vector is at parameter position 1
test_skip_normalization_for_literals: Ensures properties starting with literals are not normalized

TypeTransformer Tests (4 tests)

test_extract_type_from_property_declaration: Extracts inner type from property declarations
test_extract_type_from_vector: Extracts type from std::vector declarations
test_extract_type_from_one_or_many: Extracts type from one_or_many_property declarations
test_map_cpp_type_to_json_schema: Validates C++ to JSON Schema type mappings

BasicInfoTransformer Tests (3 tests)

test_extract_name_and_description: Validates basic property name and description extraction
test_skip_lambda_validators: Ensures lambda validator parameters are correctly skipped when finding property names
test_normalize_file_path: V

Read from source at commit 2d05341775b0OBSERVED · 2026-10-11

Exposed tools (131)

125 read · 4 write · 2 destructive. Blast radius: 2 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.

Tool	Risk	Description
`AI`	read	AI processors.
`MyConnector`	read	A simple connector.
`Parsing`	read	Parsers.
`Services`	read	Inputs that consume from services.
`acl`	read	Acl command.
`admin_api`	read	x
`agent`	read	Agent command.
`ai`	read	AI command.
`aim`	read	not ai: prefix must respect word boundary
`api-description-too-short`	read	Prose doc string too short to explain anything
`api-missing-terminal-period`	read	Prose doc string with no terminal period
`api-name-echo`	read	Description merely restates the declared name (tautology)
`api-rpc-summary-not-separated`	read	rpc comment has no blank line, so the whole comment becomes the summary
`api-summary-multiline`	read	An operation summary must be a single line
`api-summary-terminal-period`	read	An operation summary takes no terminal period
`api-undocumented-field`	read	Published field or message with no doc string
`api_version`	read	Affects endpoint paths (
`audit_overrides`	read	Audit docs-side overrides against extracted source strings
`auto_delete`	destructive	Whether the queue auto-deletes.
`balancer-status`	read	Check status. FIELDS Original fields content here. NOTES Some notes.
`bang-bang-callout`	read	!!TEXT!! style callouts are not allowed
`bar`	read	desc2
`baz`	read	no match
`blank`	read
`broken-macro`	read	xref:/glossterm:/config_ref: macro with unbalanced brackets
`bucket`	read	Bucket name.
`bundle`	read	Collect a bundle.
`checkpoint_limit`	read	The maximum number of messages that can be processed at a given time per partition.
`client_secret`	read	Sensitive: prefer environment variable interpolation (
`cloud`	read	Cloud command.
`cluster`	read	Cluster desc
`cluster_id`	read	Cluster identifier.
`coming-soon-placeholder`	read	Coming soon placeholder content
`connect`	read	Redpanda Connect plugin
`connect-availability-prose`	read	Cloud or cgo availability written as prose instead of data
`connect-cli-usage-multiline`	read	urfave/cli Usage is a one-line label
`connect-env-conditional`	read	AsciiDoc ifdef/ifndef conditional in a Go spec string
`connect-short-description-markup`	read	ShortDescription is plain text in form-based editors
`connect-xref-module`	read	xref without a module, or to a module the Connect docs do not have
`count`	read	Old desc
`create`	write	Original description
`data_dir`	read	y
`dead-marker`	read	A # -- marker no docs pipeline can attach to any key
`debug`	read	Debug a Redpanda SQL cluster.
`delete`	destructive	Delete topic
`duplicate-content`	read	Duplicate paragraphs or sections
`em-dash`	read	Em dash in prose that ships to docs.redpanda.com
`empty-description`	read	Declaration ships with no doc string at all
`fieldA`	read	original desc
`foo`	read	desc1
`generate_bundle_openapi`	read	Bundle Redpanda OpenAPI fragments. Creates complete OpenAPI 3.1 documents for admin and/or connect APIs by bundling fragments from the Redpanda repository. Writers use this when updating API reference docs.
`generate_cloud_regions`	read	Generate cloud regions table documentation. Creates a Markdown or AsciiDoc table of cloud regions and tiers from GitHub YAML data. Writers use this when updating cloud region documentation.
`generate_crd_docs`	read	Generate Kubernetes CRD (Custom Resource Definition) documentation. Creates AsciiDoc documentation for Kubernetes operator CRDs. Writers use this when updating operator reference docs.
`generate_helm_docs`	read	Generate Helm chart documentation. Creates AsciiDoc documentation for Helm charts from local directories or GitHub repositories. Writers use this when updating Helm chart reference docs.
`generate_metrics_docs`	read	Generate Redpanda metrics documentation for a specific version. Creates the public metrics reference page. Writers use this when updating metrics docs for a new release. Can run in background with progress streaming.
`generate_property_docs`	read	Generate property documentation
`generate_rpcn_connector_docs`	read	Generate Redpanda Connect connector documentation. Creates component documentation for all connectors. Writers use this when updating connector reference docs. Automatically uses docs-data/overrides.json if present.
`generate_rpk_docs`	read	Generate rpk command-line documentation using the JSON-based
`get_antora_structure`	read	Get information about the Antora documentation structure in the current repository, including components, modules, and available directories. Use this to understand the docs organization.
`get_console_version`	read	Get the latest Redpanda Console version information including version number, Docker tag, and release notes URL.
`get_job_status`	read	Get the status and progress of a background job. Use this to check on long-running documentation generation tasks.
`get_redpanda_version`	read	Get the latest Redpanda version information including version number, Docker tag, and release notes URL. Writers use this to find out what version to document.
`go-field-name-first`	read	Description leads with the Go field name instead of the YAML key
`group`	read	x
`health`	read	Health command.
`info`	read	Info command.
`inherited-type-description`	read	Field with no comment publishes its Go type\
`inline-allcaps-warning`	read	ALLCAPS warnings in middle of paragraph (not standalone admonitions)
`inner`	read	orig
`install`	write	y
`internal`	read	Partial-only child.
`invalid-xref-syntax`	read	xref: followed by http URL (invalid)
`k8s`	read	Kubernetes plugin
`latin-abbreviation`	read	Latin abbreviation instead of plain English
`lint`	read	Lint a config
`list`	read	List command.
`list_jobs`	read	List all background jobs with optional filtering. Use this to see recent documentation generation jobs and their status.
`llm`	read	x
`login`	read	Login command.
`markdown-emphasis`	read	Markdown-style text should be AsciiDoc
`misattached-marker`	read	One # -- marker is published under two different keys
`missing-field-description`	read	Config field constructed with no .Description()
`missing-inline-code`	read	Code value or field name in prose without inline code
`missing-terminal-period`	read	Property descriptions are complete sentences ending in a period
`name-echo`	read	Description merely restates the declared name (tautology)
`offset`	read	Original offset desc
`org_url`	read	Production orgs use
`otlp_http`	read	Traces go to
`p`	read	Already a string.
`partitions`	read	Original
`plain`	read	Untouched string.
`prop_redundant`	read	Same text.
`prop_upstreamable`	read	Old text.
`raw-external-urls`	read	Raw external URLs that should be links or removed
`raw-pipe`	read	Raw unescaped \| outside backticks breaks AsciiDoc tables
`replication`	read	Original too
`review_generated_docs`	read	Validate auto-generated documentation for structural issues. Checks for missing descriptions, invalid $refs, DRY violations, invalid xrefs, and calculates quality scores. For style and tone review, use the content-reviewer agent from docs-team-standards.
`rpc_server`	read	Mentions admin_api and also data_dir in one paragraph.
`rpk`	read	Root
`rpk-long-allcaps-heading`	read	ALLCAPS line in Long becomes a === heading in the docs
`rpk-long-block-delimiter`	read	A standalone ==== or ---- line in Long breaks AsciiDoc blocks
`rpk-short-multiline`	read	Short must be a single line
`rpk-terminal-period`	read	Short and flag usage strings do not end with a period
`run`	write	Run a pipeline
`run_doc_tools_command`	write	Advanced: Run a raw doc-tools command. Only use this if none of the specific tools above fit your needs. Requires knowledge of doc-tools CLI syntax.
`secret`	read	Secret command.
`security`	read	Security command.
`sentinel-default-unexplained`	read	Property defaults to a sentinel whose meaning the description never explains
`sql`	read	Interact with a Redpanda SQL cluster.
`sqlx`	read	prefix must respect word boundary
`starts-lowercase`	read	Description starts with a lowercase letter
`test`	read	Description. EXAMPLES Some examples.
`test-cmd`	read	Test command. EXAMPLES First example rpk test-cmd --flag1 Second example to exclude rpk test-cmd --exclude-me Third example rpk test-cmd --flag3
`test_property`	read	The source description.
`too-short`	read	Description is under 20 characters
`topic`	read	Topic commands
`trailing-period`	read	Metric descriptions do not end with a period
`trim-prefix`	read	Trim records
`truncated-description`	read	Description appears truncated or garbled
`unbalanced-backticks`	read	Odd number of backticks ships broken markup
`undocumented-field`	read	Exported user-facing field with no doc comment
`undocumented-key`	read	Key helm-docs publishes with an empty description
`unexpanded-jargon`	read	Internal jargon that ships to users unexpanded
`unknown-attribute`	read	{attr} reference that is not a known product attribute
`unverifiable-description`	read	Description argument is not a plain string literal
`url`	read	Queue URL.
`urls`	read	A list of URLs.
`username`	read	Username for authentication
`version`	read	Version command.
`widget`	read	Widget things.
`zero_key_warn_interval`	read	Warn interval.

Trust audit

BLOCKgrade F · trust 52/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.

Layer	What it checks	Result
L0	Provenance & inventory	WARN
L1	Static analysis of the code	FAIL
L2	Instruction surface (what it tells the agent)	PASS
L3	Class-specific surface	WARN
L4	Behavioural (sandbox)	SKIPPED

What the source does

Filesystem: declared (5 observation(s))
Network: none-observed
Shell: declared (3 observation(s))
Dependencies: not all pinned
Secrets in source: none-found

Findings (25)

HIGHCode injection · code.deserialize · CWE-78, CWE-94, CWE-95

bin/mcp-tools/antora.js:48

playbookContent = yaml.load(fs.readFileSync(playbookPath, 'utf8'));

Why it matters. deserialises untrusted bytes into live objects

Fix. use json or yaml.safe_load

HIGHCode injection · code.deserialize · CWE-78, CWE-94, CWE-95

bin/mcp-tools/antora.js:82

const content = yaml.load(fs.readFileSync(ymlPath, 'utf8'));

Why it matters. deserialises untrusted bytes into live objects

Fix. use json or yaml.safe_load

HIGHCode injection · code.eval_exec · CWE-78, CWE-94, CWE-95

tools/property-extractor/property_extractor.py:1786

result = eval(' '.join(evaluated_tokens))

Why it matters. evaluates text as code

Fix. remove; use a parser or a dispatch table

MEDIUMInventory / provenance · inv.binary · CWE-1104

docker-compose/25.1/transform/regex.wasm

regex.wasm

Why it matters. a compiled or binary member cannot be reviewed from source

Fix. ship source, or explain the binary in the README

MEDIUMInventory / provenance · inv.binary · CWE-1104

docker-compose/transform/regex.wasm

regex.wasm

Why it matters. a compiled or binary member cannot be reviewed from source

Fix. ship source, or explain the binary in the README

MEDIUMInformation disclosure · disclose.log_secret · CWE-209, CWE-532

extensions/util/connect-docs-asset.js:140

logger.warn(`The GitHub API found no ${tag} release of ${OWNER}/${REPO} with the GitHub token, and the public download has no ${ASSET_NAME} either. If the repo is private, check that the token has rea

MEDIUMFilesystem / path · mcp.destructive_tools · CWE-22, CWE-59

auto_delete, delete

Why it matters. 2 tool(s) can delete or overwrite

Fix. prefer a read-only mode or scoped tokens; the page states the blast radius

LOWInventory / provenance · inv.binary · CWE-1104

__tests__/fixtures/e2e-site/ui-bundle.zip

ui-bundle.zip

Why it matters. a compiled or binary member cannot be reviewed from source

Fix. ship source, or explain the binary in the README

LOWInventory / provenance · inv.hidden_file · CWE-1104

.release-please-manifest.json

.release-please-manifest.json

Why it matters. hidden member outside the usual dotfiles

Fix. review its purpose

LOWInventory / provenance · inv.no_license · CWE-1104

Why it matters. no LICENSE file and no repo licence

Fix. add a licence

LOWCode injection · code.deserialize · CWE-78, CWE-94, CWE-95

.github/scripts/embed-guard-script.js:51

const parsed = yaml.load(workflowSource);

Why it matters. deserialises untrusted bytes into live objects

Fix. use json or yaml.safe_load

LOWCode injection · code.deserialize · CWE-78, CWE-94, CWE-95

__tests__/workflows/kapa-source-groups-drift.test.js:99

const parsed = yaml.load(fs.readFileSync(WORKFLOW_PATH, 'utf8'));

Why it matters. deserialises untrusted bytes into live objects

Fix. use json or yaml.safe_load

LOWCode injection · code.deserialize · CWE-78, CWE-94, CWE-95

__tests__/workflows/kapa-source-groups-drift.test.js:328

const parsed = yaml.load(workflow);

Why it matters. deserialises untrusted bytes into live objects

Fix. use json or yaml.safe_load

LOWCode injection · code.eval_exec · CWE-78, CWE-94, CWE-95

__tests__/workflows/property-corpus-drift.test.js:343

const evalCond = (expr) => eval(expr.replace(/github\.ref/g, JSON.stringify(ref)));

Why it matters. evaluates text as code

Fix. remove; use a parser or a dispatch table

LOWFilesystem / path · fs.traversal · CWE-22, CWE-59

.github/scripts/derive-property-snapshot.js:32

const { normalizeIncludes } = require(path.join(__dirname, '../../tools/property-extractor/helpers/applyPropertyLinks.js'))

LOWFilesystem / path · fs.traversal · CWE-22, CWE-59

__tests__/cli-utils/big-int-json.test.js:3

const bigIntJson = require('../../cli-utils/big-int-json')

LOWFilesystem / path · fs.traversal · CWE-22, CWE-59

__tests__/cli-utils/big-int-json.test.js:101

const file = path.join(__dirname, '../../preview/redpanda-connect/docs-data/connect-latest.json')

LOWFilesystem / path · fs.traversal · CWE-22, CWE-59

__tests__/cli-utils/diff-utils.test.js:7

const { updatePropertiesJsonWithVersion, resolveDiffBaseline } = require('../../cli-utils/diff-utils.js');

LOWFilesystem / path · fs.traversal · CWE-22, CWE-59

__tests__/cli-utils/diff-utils.test.js:180

expect(() => resolveDiffBaseline(tmpDir, '../../etc/passwd')).toThrow(/attachments directory/);

LOWNetwork egress · net.beacon_words · CWE-200, CWE-319

.github/workflows/doc-strings-review.yml:632

# exfiltrate it over plaintext.

LOWNetwork egress · net.raw_ip · CWE-200, CWE-319

__tests__/e2e/link-extensions.e2e.test.js:68

const externalBase = `http://127.0.0.1:${external.port}`

LOWNetwork egress · net.raw_ip · CWE-200, CWE-319

__tests__/tools/rpk-docs/generate-rpk-docs.test.js:1542

expect(formatDescription('Point it at http://127.0.0.1:8081.')).toBe('Point it at `http://127.0.0.1:8081`.')

LOWSupply chain · supply.unpinned · CWE-829, CWE-1357

package.json

@asciidoctor/tabs, @bufbuild/buf, @modelcontextprotocol/sdk, @octokit/core, @octokit/plugin-retry, @octokit/rest, @redocly/cli, ajv

Why it matters. 36 dependency range(s) float

Fix. pin exact versions or ship a lockfile

LOWSupply chain · supply.unpinned · CWE-829, CWE-1357

tools/property-extractor/package.json

handlebars

Why it matters. 1 dependency range(s) float

Fix. pin exact versions or ship a lockfile

LOWSupply chain · supply.unpinned · CWE-829, CWE-1357

tools/property-extractor/requirements.txt

setuptools, pyyaml, requests, pytest

Why it matters. 4 requirement(s) not pinned with ==

Fix. pin exact versions

Gates applied: no_behavioural_pass, no_license.

Audited 2026-10-11 · audit v0.4.1 · source sha 2d05341775b0full audit observations/trust-audit/mcp-server/redpanda-data__docs-extensions-and-macros.json · Report an issue / request a re-scan

Audit history

Every audit this server has had. A grade with a past is a grade somebody is still checking.

Date	Source	Verdict	Grade	Score	Change
2026-10-11	`2d05341775b0`	BLOCK	F	52	first audit

Questions

What is the @redpanda-data/docs-extensions-and-macros MCP server?

Extensions and macros developed for Redpanda documentation.

What tools does @redpanda-data/docs-extensions-and-macros expose?

131 in total: 125 read-only, 4 that write, and 2 that can delete or overwrite (auto_delete, delete). Every one is listed on this page with its risk.

Is @redpanda-data/docs-extensions-and-macros safe to connect to an agent?

No — not without reading the findings first. The audit graded it F (52/100) and found 3 critical or high issues in the source. Each one is listed on this page with the file and line it is on. Separately from the audit: 2 of its tools can destroy data, so scope the token you give it to what you actually need.

What credentials does @redpanda-data/docs-extensions-and-macros need?

It reads ACTIONS_BOT_TOKEN, ALGOLIA_ADMIN_API_KEY, GH_TOKEN, GITHUB_TOKEN, GIT_CREDENTIALS and REDPANDA_GITHUB_TOKEN from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.

How does @redpanda-data/docs-extensions-and-macros run?

It speaks stdio, so it runs as a local process your client starts.

How current is this page?

The grade is for one exact copy of the source (2d05341775b0), read on 2026-10-11. The repository is watched and re-audited when it changes.

Source	redpanda-data/docs-extensions-and-macros
Transports	stdio
Credentials it reads	`ACTIONS_BOT_TOKENALGOLIA_ADMIN_API_KEYGH_TOKENGITHUB_TOKENGIT_CREDENTIALSREDPANDA_GITHUB_TOKEN`
Activity	pushed 0d ago