Repomap Query Claims

When to use this skill

Use this skill when you need to:

•Understand what architectural patterns exist in a module
•Check if a module uses a specific pattern (facade, repository, strategy, etc.)
•Assess confidence in architectural claims before refactoring
•Find modules with high boil risk (needing re-verification)
•Understand dependencies before making changes

Prerequisites

•Canonical claims file exists: repomap_claims.jsonl
•Claims are up-to-date (run /claims-pipeline.md if needed)

Query Patterns

Query by Module

Find all claims about a specific module:

bash

grep '"repomap.claims.langchain"' repomap_claims.jsonl | head -5

Or use jq for structured output:

bash

cat repomap_claims.jsonl | jq 'select(.statement | contains("langchain")) | {claim_id, statement, confidence, claim_tier}'

Query by Claim Type

Find all claims of a specific type:

bash

cat repomap_claims.jsonl | jq 'select(.claim_type == "seam_boundary") | {claim_id, statement, confidence}'

Claim types:

•seam_boundary - Architectural boundaries and module interfaces
•integration_touchpoint - External system integration points
•intent_glossary - Domain concepts and terminology
•layer_violation - Violations of layered architecture

Query by Signed Confidence

Find strongly supported claims (> +0.7):

bash

cat repomap_claims.jsonl | jq 'select(.confidence > 0.7) | {claim_id, statement, confidence}'

Find contradicted claims (< -0.3):

bash

cat repomap_claims.jsonl | jq 'select(.confidence < -0.3) | {claim_id, statement, confidence, boil_risk}'

Find uncertain claims (near zero):

bash

cat repomap_claims.jsonl | jq 'select(.confidence > -0.3 and .confidence < 0.3) | {claim_id, statement, confidence, boil_risk}'

Query by Boil Risk

Find claims with high boil risk (> 0.1):

bash

cat repomap_claims.jsonl | jq 'select(.boil_risk > 0.1) | {claim_id, statement, confidence, boil_risk, pressure, gravity}'

Query by Pattern

Find all facade pattern claims:

bash

cat repomap_claims.jsonl | jq 'select(.operational_definition_ref == "facade_pattern:0.1") | {claim_id, statement, confidence}'

Pattern refs:

•facade_pattern:0.1
•repository_pattern:0.1
•strategy_pattern_opportunity:0.1
•data_clump:0.1
•layered_architecture:0.1

Interpreting Results

Signed Confidence Scores

Positive (Supporting Evidence):

•+0.7 to +1.0 - Strong supporting evidence
•+0.3 to +0.7 - Moderate supporting evidence
•+0.0 to +0.3 - Weak supporting evidence

Negative (Contradictory Evidence):

•-0.3 to -0.0 - Weak contradictory evidence
•-0.7 to -0.3 - Moderate contradictory evidence
•-1.0 to -0.7 - Strong contradictory evidence

Near Zero:

•-0.3 to +0.3 - Uncertain, no clear evidence either way

Boil Risk

•> 0.2 - Critical, needs immediate re-verification
•0.1-0.2 - High risk, should re-verify soon
•0.05-0.1 - Moderate risk, monitor
•< 0.05 - Low risk, stable

Thermodynamic Properties

•Pressure - Urgency from code churn (higher = more changes)
•Gravity - Importance weight (higher = more critical)
•Temperature - Rate of confidence change

Example Workflow

Before Refactoring a Module

•

Query claims about the module:

bash

cat repomap_claims.jsonl | jq 'select(.statement | contains("langchain.orchestrator")) | {claim_id, statement, confidence, boil_risk}'

•

Check dependencies:

bash

grep "repomap.claims.langchain.orchestrator" .repomap/deps.edgelist

•
Assess risk:
- •High confidence + low boil risk = safe to refactor
- •Low confidence + high boil risk = re-verify first
•
Make informed decision based on evidence

Integration with Other Skills

•repomap-claims-ops - Run pipeline to update claims
•repomap-codebase-retrieval - Semantic search for code
•sequential-thinking-default - Plan refactoring strategy

References

•Canonical claims: repomap_claims.jsonl
•Artifacts: .repomap/
•Workflow: /claims-pipeline.md