Skip to main content
The MCP server exposes 5 tools for interacting with the traceability graph. All tools return JSON and include next-step hints.

​
sdd_query

Search artifacts by text, ID, type, or domain. Returns matching artifacts with relevance scores.
​
query
string
required
Search text — matches against artifact ID, title, file path, and code/test refs
​
type
string
Filter by artifact type: REQ, UC, WF, API, BDD, INV, ADR, NFR, RN, FASE, TASK
​
domain
string
Filter by business domain (e.g. ‘Security & Auth’, ‘Frontend & UI’)
​
limit
number
default:"20"
Max results to return

​
Example

{
  "query": "authentication",
  "type": "REQ",
  "limit": 10
}
Returns artifacts ranked by relevance score (exact ID match = 100, partial ID = 60, title match = 40, code ref = 30).

​
sdd_trace

Trace the complete traceability chain for an artifact: REQ → UC → WF → API → BDD → INV → ADR → TASK → COMMIT → CODE → TEST. Identifies breaks in the chain and reports completeness status.
​
artifact_id
string
required
The artifact ID to trace (e.g. REQ-AUTH-001, UC-003)

​
Example

{
  "artifact_id": "REQ-AUTH-001"
}
Returns a chain organized by type level, with status:
  • COMPLETE: No breaks in expected chain
  • PARTIAL: 1-2 missing links
  • FRAGMENTED: 3+ missing links

​
sdd_impact

Analyze blast radius of changing an artifact. Uses BFS by depth to classify impact.
​
artifact_id
string
required
The artifact ID to analyze (e.g. REQ-AUTH-001, UC-003)
​
direction
string
required
Direction to traverse:
  • upstream — what does this depend on
  • downstream — what depends on this
​
maxDepth
number
default:"3"
Maximum traversal depth

​
Impact classification by depth

  • Depth 1: WILL_BREAK — direct dependents
  • Depth 2: LIKELY_AFFECTED — indirect dependents
  • Depth 3: MAY_NEED_REVIEW — transitive dependents

​
Risk levels

  • HIGH: 5+ depth-1 artifacts OR 20+ total affected
  • MEDIUM: 2-5 depth-1 artifacts OR 10-20 total affected
  • LOW: <2 depth-1 artifacts OR <10 total affected

​
Code intelligence enrichment

When code intelligence is available (via /sdd:code-index), the tool includes:
  • Direct symbols linked to the artifact
  • Transitive callers from the call graph
  • Total call chain depth

​
sdd_coverage

Analyze traceability coverage gaps grouped by business domain and technical layer. Identifies uncovered requirements and top priority gaps.
​
domain
string
Filter by business domain (e.g. ‘Security & Auth’)
​
layer
string
Filter by technical layer: Infrastructure, Backend, Frontend, Integration/Deployment

​
Coverage criteria

For each REQ, checks for:
  • UC link: Use case implements the requirement
  • BDD link: BDD scenario verifies the requirement
  • Code refs: Implementation symbols
  • Test refs: Test cases

​
Coverage status

  • Covered: All 4 criteria met (0 missing)
  • Partial: 1-3 criteria missing
  • Uncovered: All 4 criteria missing

​
Code intelligence enrichment

When available, includes:
  • Total symbols in the codebase
  • Symbols with explicit artifact annotations
  • Symbols with inferred references
  • Annotated and total coverage percentages

​
sdd_context

Get a 360-degree view of an artifact: its definition, all upstream/downstream connections, code references, test references, commit references, and coverage gaps. Essential before modifying any artifact.
​
artifact_id
string
required
The artifact ID (e.g. REQ-AUTH-001, UC-003, TASK-F01-003)

​
Example

{
  "artifact_id": "REQ-AUTH-001"
}

​
Coverage status classifications

  • Complete: Has UC + BDD + code + tests
  • In Progress: Has UC + (code OR tests)
  • Specified: Has UC only
  • Not Started: No UC link

​
Gap detection for REQ artifacts

The tool identifies specific gaps:
  • ORPHAN: No relationships found
  • MISSING_UC: No use case implements this requirement
  • MISSING_BDD: No BDD scenario verifies this requirement
  • MISSING_TASK: No task decomposes this requirement
  • MISSING_CODE: No code references found
  • MISSING_TESTS: No test references found
  • MISSING_COMMITS: No commit references found

​
Code intelligence enrichment

When available, includes:
  • Symbols linked to the artifact (explicit + inferred)
  • Symbol type, file path, line range
  • Callers and callees from the call graph
  • Associated processes (if the artifact is part of a workflow)

Build docs developers (and LLMs) love

Get started for freeTalk to us