mountain-nomad-bc/pacify-x/dependency-impact-tracer

dependency-impact-tracer

Purpose

Predict downstream effects of changing a skill, tool, schema, state, model, policy, or data contract.

Use when

• component change proposed
• deprecation or migration planned

Do not use when

• documentation-only change has no executable effect

Required inputs

• behavior graph
• dependency graph
• registry refs
• tests
• telemetry consumers
• versions

Outputs

• affected components
• required tests
• migration steps
• risk
• rollback dependencies

Procedure

1. locate changed node.
2. traverse typed dependencies.
3. separate code/runtime/data/policy effects.
4. map tests/dashboards.
5. rank consequence/confidence.
6. emit changed-area validation selectors covering owners, direct and transitive consumers, contracts, migrations, routes, persistence, deployment assets, and prior failure clusters.
7. keep unknown dynamic edges in the mandatory review denominator; absence from the graph is not proof of no impact.

Guardrails

• not all edges are equal.
• include trace-discovered consumers.
• flag unknown dynamic links.

Integration points

• behavior mapper
• registries
• CI gates
• promotion

Failure modes to test

• incomplete graph
• stale registry
• unmapped dynamic integration

Operational metrics

• escaped impact defects
• false positives
• test selection accuracy

Completion requirements

The skill is complete only when structured outputs are emitted, evidence references resolve, declared postconditions are checked, and the execution wrapper records the result. Any memory remains a candidate until framework promotion controls accept it.

Research basis

• Harness Handbook

Runtime binding

• Family: impact
• Binding: runtime.operational_controls.run_control
• Activation: metadata is discoverable at startup; load this body only after selection.
• Effects: read-only control-plane analysis unless a separately admitted composed capability declares more.

Completion and evidence

Return a structured decision, reasons, outputs, and evidence references. Treat unresolved provenance, permissions, required inputs, or postconditions as a fail-closed result. Candidate memory, research, generated skills, and speculative work never become active without separate admission and evidence.