By RayforceDB
Architectural telemetry for AI coding agents. Phase-scoped skills over the raysense MCP server: scan + baseline at session start, blast-radius before edits, regression diff after, on-demand architecture audits, time-window drift detection.
Use when the user explicitly asks for a structural audit, architecture review, dead-code report, test-gap analysis, or evolution hotspot scan. Heavier and noisier than the other raysense skills — only run on demand, not as part of the routine edit loop.
Use at the start of any coding session in a repository to scan the structure, save a baseline, and materialize splayed-table memory for fast follow-up queries. Establishes the "before" reference point that the verify skill diffs against later.
Use after a rescan to surface structural regressions across a time window. Diffs the latest scan against the saved baseline AND the trend history, ranking dimensions that worsened, files newly hot, and rules newly tripped. Configurable window (7d, 30d, 90d). Use periodically (daily, weekly, or pre-PR).
Use before refactoring, deleting, moving, or substantially modifying a file to compute its blast radius, coupling profile, and cycle exposure. Lets the agent edit with awareness of downstream effects rather than discovering them after the fact.
Use when the agent has a structural question about the saved baseline that the typed MCP tools (health, hotspots, rules, blast radius, coupling, cycles, evolution) do not directly answer. Runs Rayfall expressions against splayed baseline tables via raysense_baseline_query. Three modes are available - select queries for filter/project/aggregate (most common), .graph.* algorithms (PageRank, Louvain, topsort, shortest-path, betweenness, closeness, MST, k-shortest, BFS expand) for centrality and reachability over the call graph, and Datalog rules with transitive closure for declarative reachability ("reaches", "depends-on", "tainted-by"). Reach for this when the question shape is "files where X and Y," "most-central callers," "what does X transitively reach," or any custom slice across the 18 baseline tables.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Raysense reads your repository as a graph: who imports who, where the cycles are, which files are now load-bearing, what tends to change together. It runs locally, refreshes on save, and serves the whole picture to your coding agent over MCP. Before an edit, the agent can ask what depends on this file. After a chunk of edits, it can ask did this regress anything.
A coding agent reads source one file at a time. The shape of the project (its modules, its layers, its cycles, the files that always change together) never reaches its working memory. Reviewers operate on diffs, and a diff hides structure by definition. So architectural drift is invisible until it shows up as a production bug, a regression, or a refactor that takes a week.
Six dimensions, each graded A through F against the dependency graph and commit history of the repo. The overall score, 0 to 100, is their weighted aggregate:
The score moves with structure, not with cosmetics: adding tests or shuffling files around will not lift it.
cargo install raysense
raysense . # health report
raysense . --check # CI gate, exits non-zero on rule violations
raysense . --watch # rescan + reprint on a 2s loop
raysense . --ui # live dashboard at http://localhost:7000
raysense --mcp # stdio MCP server for agents
Pointed at this very repo (raysense .):
score 82 / 100
coverage 90 / 100
structure 68 / 100
facts files=34 functions=656 calls=7518 call_edges=1383 imports=247
imports local=98 external=124 system=0 unresolved=25
graph resolved_edges=89 cycles=0 max_fan_in=53 max_fan_out=21
coupling local_edges=98 cross_module_edges=0 god_files=2 unstable_hotspots=0
size max_file_lines=5907 max_function_lines=1345 large_files=7 long_functions=20
test_gap production_files=13 test_files=0 files_without_nearby_tests=13
dimensions modularity=100/100 (A) acyclicity=100/100 (A) depth=100/100 (A)
equality=45/100 (F) redundancy=80/100 (B) structural_uniformity=79/100 (C)
overall_grade B
architecture depth=4 max_blast_radius=7 max_blast_radius_file=src/facts.rs
complexity max=140 avg=4.261 cognitive_max=119 cognitive_avg=3.457 dead_functions=50
evolution available=true commits_sampled=151 changed_files=34 authors=2 bug_fix_commits=1
npx claudepluginhub rayforcedb/raysense --plugin raysenseComprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Develop, test, build, and deploy Godot 4.x games with Claude Code. Includes GdUnit4 testing, web/desktop exports, CI/CD pipelines, and deployment to Vercel/GitHub Pages/itch.io.
Access thousands of AI prompts and skills directly in your AI coding assistant. Search prompts, discover skills, save your own, and improve prompts with AI.
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer
Intelligent draw.io diagramming plugin with AI-powered diagram generation, multi-platform embedding (GitHub, Confluence, Azure DevOps, Notion, Teams, Harness), conditional formatting, live data binding, and MCP server integration for programmatic diagram creation and management.
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review