Functions — github.com/Exorust/Adkit-MCP

hub / github.com/Exorust/Adkit-MCP / functions

Functions155 in github.com/Exorust/Adkit-MCP

⨍ Functions 155◇ Types & classes 46

↓ 24 callersMethodmatch

(self, request: MatchRequest)

src/ad_injector/services/match_service.py:50

↓ 22 callersFunction_simple_request

(**overrides)

tests/test_match_service.py:106

↓ 21 callersFunction_build_service

(hits: list[VectorHit] | None = None)

tests/test_match_service.py:95

↓ 17 callersFunction_make_hit

Build a VectorHit matching the port's return type.

tests/test_match_service.py:35

↓ 11 callersMethod_get_client

(self)

src/ad_injector/adapters/qdrant_vector_store.py:39

↓ 10 callersFunction_make_hit

( ad_id: str, score: float, *, sensitive: bool = False, age_restricted: bool = False,

tests/test_policy_engine.py:10

↓ 8 callersFunction_build_filter

(**constraint_kwargs)

tests/test_targeting_engine.py:10

↓ 8 callersMethodapply

Return only hits that pass all policy checks.

src/ad_injector/domain/policy_engine.py:31

↓ 8 callersFunctionget_settings

Return the singleton RuntimeSettings (cached after first call).

src/ad_injector/config/runtime.py:70

↓ 8 callersFunctionrequire_admin_scope

Require admin scope for Control Plane. Raises PermissionError if not allowed.

src/ad_injector/mcp/auth.py:8

↓ 7 callersFunction_get_index_service

()

src/ad_injector/mcp/tools.py:85

↓ 6 callersFunctioncreate_server

Build and return a configured FastMCP server. Args: mode: ``"data"`` for the Data Plane (LLM-facing, read-only) or ``"admin

src/ad_injector/mcp/server.py:19

↓ 3 callersMethod_ad_id_to_uuid

(self, ad_id: str)

src/ad_injector/adapters/qdrant_vector_store.py:48

↓ 3 callersFunction_get_tool_names

Extract registered tool names from a FastMCP server.

tests/test_data_plane_tools.py:29

↓ 3 callersFunctionbuild_index_service

Construct an IndexService with real adapters.

src/ad_injector/wiring.py:26

↓ 3 callersMethodcollection_info

(self)

src/ad_injector/ports/vector_store.py:40

↓ 2 callersMethod_blocked_keywords_intersect

True if ad.blocked_keywords intersects context_text tokens (substring or exact).

src/ad_injector/domain/policy_engine.py:77

↓ 2 callersMethod_get_model

(self)

src/ad_injector/adapters/fastembed_provider.py:17

↓ 2 callersMethod_translate_condition

(f: FieldFilter)

src/ad_injector/adapters/qdrant_vector_store.py:286

↓ 2 callersMethodbuild_filter

( self, constraints: MatchConstraints, placement: PlacementContext, )

src/ad_injector/domain/targeting_engine.py:22

↓ 2 callersFunctionbuild_match_service

Construct a MatchService with real adapters.

src/ad_injector/wiring.py:17

↓ 2 callersFunctioncheck_scope

Check scope for the given server mode. Call at server start or per-request.

src/ad_injector/mcp/auth.py:30

↓ 2 callersMethodembed

(self, text: str)

src/ad_injector/ports/embedding.py:12

↓ 2 callersMethodembed

(self, text: str)

src/ad_injector/adapters/fastembed_provider.py:22

↓ 2 callersMethodensure_collection

(self, dimension: int)

src/ad_injector/ports/vector_store.py:36

↓ 2 callersMethodupsert_ads

(self, ads: list[Ad])

src/ad_injector/services/index_service.py:50

↓ 1 callersMethod_allowed

( self, hit: VectorHit, constraints: MatchConstraints, context_text: str,

src/ad_injector/domain/policy_engine.py:62

↓ 1 callersMethod_ensure_enabled_filter

Merge in must_not enabled=False so only enabled ads are returned.

src/ad_injector/adapters/qdrant_vector_store.py:256

↓ 1 callersMethod_filter_spec_to_qdrant

Convert simple dict filter (e.g. advertiser_id, ad_id) to Qdrant Filter.

src/ad_injector/adapters/qdrant_vector_store.py:264

↓ 1 callersMethod_get_collection_meta

(self)

src/ad_injector/adapters/qdrant_vector_store.py:142

↓ 1 callersFunction_get_match_service

()

src/ad_injector/mcp/tools.py:80

↓ 1 callersMethod_hit_to_candidate

(self, hit: VectorHit, request_id: str)

src/ad_injector/services/match_service.py:135

↓ 1 callersMethod_set_collection_meta

( self, dimension: int, embedding_model_id: str, schema_version: str, )

src/ad_injector/adapters/qdrant_vector_store.py:159

↓ 1 callersFunction_shape_ads_get

(payload: dict | None)

src/ad_injector/mcp/tools.py:61

↓ 1 callersFunction_shape_collection_ensure

(d: dict)

src/ad_injector/mcp/tools.py:57

↓ 1 callersFunction_shape_collection_info

(d: dict)

src/ad_injector/mcp/tools.py:53

↓ 1 callersFunction_shape_match_response

Return only allowed fields for ads.match response.

src/ad_injector/mcp/tools.py:41

↓ 1 callersFunction_store_trace_for_explain

Store audit trace keyed by each match_id for ads.explain.

src/ad_injector/mcp/tools.py:67

↓ 1 callersFunction_tokenize_context

Deterministic tokenization: split on whitespace, lowercase.

src/ad_injector/domain/policy_engine.py:17

↓ 1 callersMethod_translate_filter

(vf: VectorFilter)

src/ad_injector/adapters/qdrant_vector_store.py:277

↓ 1 callersMethodbulk_disable

(self, filter_spec: dict)

src/ad_injector/ports/vector_store.py:50

↓ 1 callersMethoddelete_ad

(self, ad_id: str)

src/ad_injector/ports/vector_store.py:46

↓ 1 callersMethoddelete_collection

(self)

src/ad_injector/ports/vector_store.py:38

↓ 1 callersMethodget_ad

(self, ad_id: str)

src/ad_injector/ports/vector_store.py:48

↓ 1 callersFunctionload_ads_from_file

Load ads from a JSON file. Raises on missing file or invalid JSON/schema.

src/ad_injector/cli.py:15

↓ 1 callersFunctionlog_tool_invocation

Emit structured log and update metrics stub.

src/ad_injector/mcp/observability.py:19

↓ 1 callersFunctionmain

()

src/ad_injector/cli.py:46

↓ 1 callersFunctionmain

()

src/ad_injector/main_control.py:18

↓ 1 callersFunctionmain

Main function for the ad-injector MCP server.

src/ad_injector/main.py:6

↓ 1 callersFunctionmain

()

src/ad_injector/main_runtime.py:18

↓ 1 callersMethodnew_match_id

(self, request_id: str, ad_id: str)

src/ad_injector/ports/id_gen.py:20

↓ 1 callersMethodnew_request_id

(self)

src/ad_injector/ports/id_gen.py:13

↓ 1 callersMethodquery

( self, vector: list[float], vector_filter: VectorFilter, top_k: int, )

src/ad_injector/ports/vector_store.py:27

↓ 1 callersMethodreason

Return audit reason for this hit: 'allowed' or 'denied: <reason>'.

src/ad_injector/domain/policy_engine.py:45

↓ 1 callersFunctionregister_control_plane_tools

src/ad_injector/mcp/tools.py:201

↓ 1 callersFunctionregister_data_plane_tools

src/ad_injector/mcp/tools.py:96

↓ 1 callersFunctionrequire_data_scope

Require data (read-only) scope for Data Plane. Raises PermissionError if not allowed.

src/ad_injector/mcp/auth.py:19

↓ 1 callersFunctionrun_server

Run the MCP server using stdio transport.

src/ad_injector/mcp_server.py:63

↓ 1 callersFunctionrun_smoke_check

Run a minimal health check (Qdrant + embedding). Returns: Dict with keys: ok (bool), qdrant (str), embedding (str), error (str | None).

src/ad_injector/ops/smoke_check.py:6

↓ 1 callersFunctionseed_ads

Load demo ads from a JSON file and upsert them via IndexService.

src/ad_injector/cli.py:36

↓ 1 callersMethodto_pinecone_metadata

Convert ad to Pinecone metadata format. Includes enabled for bulk_disable support.

src/ad_injector/models/ad.py:40

↓ 1 callersMethodupsert_batch

( self, ads_with_embeddings: list[tuple[object, list[float]]] )

src/ad_injector/ports/vector_store.py:42

Method__init__

(self, model_id: str = "BAAI/bge-small-en-v1.5")

src/ad_injector/adapters/fastembed_provider.py:13

Method__init__

(self, settings: RuntimeSettings)

src/ad_injector/adapters/qdrant_vector_store.py:27

Method__init__

( self, embedding_provider: EmbeddingProvider, vector_store: VectorStorePort,

src/ad_injector/services/index_service.py:20

Method__init__

( self, embedding_provider: EmbeddingProvider, vector_store: VectorStorePort,

src/ad_injector/services/match_service.py:32

Method__init__

(self, hits: list[VectorHit] | None = None)

tests/test_match_service.py:70

Method_collection

(self)

src/ad_injector/adapters/qdrant_vector_store.py:36

Method_port_range

(cls, v: int)

src/ad_injector/config/runtime.py:63

Functionads_bulk_disable

Set enabled=false for all ads matching the filter. Args: filter_json: JSON object e.g. {"advertiser_id": "x"} or {"ad_id": ["a","

src/ad_injector/mcp/tools.py:306

Functionads_capabilities

Supported placements, constraint keys, embedding model, schema version.

src/ad_injector/mcp/tools.py:179

Functionads_delete

Delete a single ad by ID. Args: ad_id: The ad identifier to delete Returns: JSON confirmation

src/ad_injector/mcp/tools.py:291

Functionads_explain

Return audit trace for a prior match (why eligible/ineligible, filters, scores). Args: match_id: Opaque ID returned with each can

src/ad_injector/mcp/tools.py:154

Functionads_get

Get a single ad by ID (debugging). Args: ad_id: The ad identifier Returns: JSON with ad payload (allowlisted

src/ad_injector/mcp/tools.py:327

Functionads_health

Liveness/readiness: Qdrant and embedding provider reachable.

src/ad_injector/mcp/tools.py:169

Functionads_match

Match ads by text query. Returns similar ads based on semantic similarity. Args: query: Text query to match against ads top_k: Nu

src/ad_injector/mcp_server.py:19

Functionads_match

Match ads by semantic context (read-only). Returns ranked candidates and match_id for explain. Args: context_text: Conversational

src/ad_injector/mcp/tools.py:100

Functionads_upsert_batch

Upsert a batch of ads. Validate Ad schema, embed, upsert. Size limit from config. Args: ads_json: JSON array of ad objects

src/ad_injector/mcp/tools.py:263

Methodbulk_disable

Set enabled=False for all points matching filter_spec. Returns count updated.

src/ad_injector/adapters/qdrant_vector_store.py:219

Methodbulk_disable

Set enabled=False for all ads matching filter_spec. Returns count updated.

src/ad_injector/services/index_service.py:68

Functioncollection_ensure

Ensure the ads collection exists with the given config. Args: dimension: Embedding vector dimension embedding_model_i

src/ad_injector/mcp/tools.py:205

Functioncollection_info

Return metadata about the current ads collection. Returns: JSON with name, points_count, status, dimension, embedding_model_id, s

src/ad_injector/mcp/tools.py:231

Methodcollection_info

(self)

src/ad_injector/adapters/qdrant_vector_store.py:129

Methodcollection_info

(self)

src/ad_injector/services/index_service.py:47

Methodcollection_info

(self)

tests/test_match_service.py:85

Functioncollection_migrate

Optional: run schema migrations / re-index between versions. Args: from_version: Current schema version to_version: T

src/ad_injector/mcp/tools.py:243

Methoddelete_ad

(self, ad_id: str)

src/ad_injector/adapters/qdrant_vector_store.py:203

Methoddelete_ad

(self, ad_id: str)

src/ad_injector/services/index_service.py:61

Methoddelete_ad

(self, ad_id)

tests/test_match_service.py:87

Methoddelete_collection

(self)

src/ad_injector/adapters/qdrant_vector_store.py:123

Methoddelete_collection

(self)

src/ad_injector/services/index_service.py:44

Methoddelete_collection

(self)

tests/test_match_service.py:84

Methodembed

(self, text: str)

tests/test_match_service.py:31

Methodembed_batch

(self, texts: list[str])

src/ad_injector/adapters/fastembed_provider.py:27

Methodembedding_text

Generate the text to be embedded (title + body + topics).

src/ad_injector/models/ad.py:35

Methodensure_collection

( self, dimension: int, embedding_model_id: str | None = None, schema_version:

src/ad_injector/adapters/qdrant_vector_store.py:90

Methodensure_collection

( self, dimension: int | None = None, embedding_model_id: str | None = None, s

src/ad_injector/services/index_service.py:30

Methodensure_collection

(self, dimension)

tests/test_match_service.py:83

Methodget_ad

(self, ad_id: str)

src/ad_injector/adapters/qdrant_vector_store.py:209

Methodget_ad

Return raw ad payload (flat dict from store) or None. For MCP/CLI use.

src/ad_injector/services/index_service.py:64

next →1–100 of 155, ranked by callers