From vcsdd
Performs BFS forward-impact propagation on changed spec nodes in CEG to identify downstream docs needing updates. Auto-detects via git diff; classifies Green/Amber/Gray bands with convention alerts.
How this skill is triggered — by the user, by Claude, or both
Slash command
/vcsdd:vcsdd-coherence-impactThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Given a set of changed or affected spec nodes, perform BFS forward-impact
Given a set of changed or affected spec nodes, perform BFS forward-impact
propagation through the CEG to identify all downstream documents that may
need updating. Also surface any CoDD conventions: / must_review alerts
attached to the changed node or its immediate parents. Results are classified
into Green / Amber / Gray bands.
If no node is provided, auto-detect changed files from git diff --name-only HEAD
and resolve them to start nodes using CoDD-compatible rules:
coherence: or codd: frontmatter -> node_idfile:<path> node -> file:<path>path -> that nodecoherence.json exists (run /vcsdd-coherence-scan first)Determine start nodes in this order:
node_id arguments, use them.--diff <target>, auto-detect from that git diff target.HEAD).node_id, you may use that directly.Load the CEG and resolve changed files when auto-detecting:
const {
loadCoherence,
detectChangedNodes,
propagateImpact,
collectConventionAlerts,
generateImpactReport,
} = require('./scripts/lib/vcsdd-coherence');
const featureName = /* active feature */;
const ceg = loadCoherence(featureName);
if (!ceg) { /* coherence not active, skip */ }
const detected = detectChangedNodes(featureName, { diffTarget: 'HEAD' });
const startNodes = detected.startNodes.map(entry => entry.nodeId);
const impacts = propagateImpact(ceg, startNodes, 10, 0);
const conventionAlerts = collectConventionAlerts(ceg, startNodes);
const report = generateImpactReport(impacts, ceg, undefined, { conventionAlerts });
Present the Markdown report to the user, including:
If auto-detection returns no start nodes:
node_id values or add coherence: / codd: / source_files: metadata| Band | Confidence | Action |
|---|---|---|
| 🟢 Green | ≥90%, ≥2 evidence | Auto-propagate changes to these specs |
| 🟡 Amber | ≥50% | Require human review before updating |
| ⚪ Gray | <50% | Informational — note but do not automatically update |
## Coherence Impact Report
### 🟢 Green — Auto-propagate safe
- **System Design** (`design:system-design`) — depth 1, confidence 90%
- **Database Design** (`design:db-design`) — depth 2, confidence 81%
### 🟡 Amber — Human review required
- **Test Strategy** (`design:test-strategy`) — depth 2, confidence 72%
### ⚪ Gray — Informational only
- **Implementation Plan** (`design:impl-plan`) — depth 3, confidence 34%
When /vcsdd-feedback routes a finding to Phase 1a, automatically run this
skill and append the impact report to the history event. This enriches the
feedback context with dependency-aware information.
npx claudepluginhub sc30gsw/vcsdd-claude-code --plugin vcsddRebuilds Conditioned Evidence Graph (CEG) from coherence: or codd: frontmatter in Markdown spec files. Tracks spec-implementation dependencies for current change-impact analysis.
Performs graph-based impact analysis to answer 'if I change X, what breaks?' before merging PRs or planning refactors.
Analyzes impact of changes to sphinx-needs requirements, specs, or items by tracing links, extra_links, and codelinks to affected needs and code files.