[avrabe]

Why

#956 items 1–2. CoverageEntry::percentage() returned 100 for an empty scope, and its doc comment said display code MUST check is_empty_scope() first. Only the CLI did. The dashboard, static export, JSON API and MCP painted a green 100% for a project whose sources: loaded nothing. None of them applied coverage.unmodelled-rules either, so a project got one answer from rivet coverage and another everywhere else.

What

• API-compatible: main is already 0.40.0, so a public break fails the semver gate. percentage_opt(), accounted_percentage_opt(), overall_coverage_opt() and ClosureEntry::percentage_opt() return None for an empty scope. The 100-returning methods remain but are #[deprecated], so clippy -D warnings rejects any rivet use of them. The compiler forced every surface over.
• One rendering: coverage::format_percentage / coverage_band render None as n/a / null, in a neutral colour, never green.
• One definition of project coverage: coverage::compute_project_coverage applies the unmodelled-rules declarations. The serve API, the dashboard coverage and stats pages, and MCP use it and carry the declared reason.
• Same bug class found and fixed in three more places:
  • the export matrix's row coverage (100% for an empty row);
  • rivet context, which told an AI agent "Overall: 100.0%" and "All rules at 100% coverage" for a project with nothing loaded;
  • --fail-under, whose JSON said "passed": true in the run that exited 1 for an empty load.
• Kani: the proofs now prove None exactly when total is 0, and bounds otherwise.

Verification

• One fixture (empty scope plus a declared unmodelled rule) is asserted across the CLI (coverage JSON, the gate, rivet context), the serve API and dashboard, and MCP. Six negative controls (reintroducing 100 on each surface, dropping the reason, the gate) each fail a named test.
• Before this, the whole suite passed; no test covered what these surfaces showed for an empty scope.
• fmt, clippy 1.97 --all-targets -D warnings, cargo test --workspace --no-fail-fast (2638), rivet validate, rivet docs check, yamllint.

Residual, bound by the public API (recorded on REQ-387)

• The snapshot format stores an f64, so it keeps 100 for an empty scope.
• Core's single-page export shows n/a but not the unmodelled reason.
• Item 3 (MCP validate parity) and the lower tier remain.

Refs #956.

🤖 Generated with Claude Code

https://claude.ai/code/session_015HMQUV3u86jN2hmCtXNTc9

[github-actions]

📐 Rivet artifact delta

Change	Count
Added	0
Removed	0
Modified	1
Downstream impacted (depth ≤ 5)	0

Graph

graph LR
  REQ_387["REQ-387"]:::modified
  classDef added fill:#d4edda,stroke:#28a745,color:#155724
  classDef removed fill:#f8d7da,stroke:#dc3545,color:#721c24
  classDef modified fill:#fff3cd,stroke:#ffc107,color:#856404
  classDef overflow fill:#e2e3e5,stroke:#6c757d,color:#495057,stroke-dasharray: 3 3

Loading

Modified

ID	Changes
REQ-387

📎 Full HTML dashboard attached as workflow artifact rivet-delta-pr-1019 — download from the workflow run.

_{Posted by rivet-delta workflow. The graph shows only changed artifacts; open the HTML dashboard (above) for full context.}

[github-actions]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Rivet Criterion Benchmarks'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.20.

Benchmark suite	Current: fcd2665	Previous: d2c7c37	Ratio
store_insert/10000	20260990 ns/iter (± 1026131)	12998695 ns/iter (± 636331)	1.56
link_graph_build/10000	33752074 ns/iter (± 1306418)	24773408 ns/iter (± 1563623)	1.36
diff/10000	9405194 ns/iter (± 604730)	7659528 ns/iter (± 249548)	1.23

This comment was automatically generated by workflow using github-action-benchmark.

[codecov]

Codecov Report

❌ Patch coverage is 94.97717% with 11 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
rivet-core/src/coverage.rs	90.19%	10 Missing ⚠️
rivet-core/src/snapshot.rs	50.00%	1 Missing ⚠️

📢 Thoughts on this report? Let us know!

[avrabe]

PR-diff mutation on cdfe334 reached the job's 45-minute limit after 73 of 90 mutants: 56 caught, 17 unviable, 0 missed. I ran the 17 it didn't reach locally, in a clean worktree with no shared target dir. Nine survived, all in export.rs: the matrix row percentage and its empty-row guard, render_section_index and render_section_coverage. 7b7a026 adds two in-module tests; the rerun gives 18 caught, 1 unviable, 0 missed, and none of the 19 builds reused a stale binary. Kani failed twice with exit 143 (runner shutdown mid-compile). The harness compile error fixed in cdfe334 was real; cargo kani --harness proof_coverage_percentage_bounds verifies locally with 0.67.0.

🤖 Generated with Claude Code

https://claude.ai/code/session_015HMQUV3u86jN2hmCtXNTc9