Coverage

Test Coverage Map

> Purpose: Query test coverage, identify uncovered areas, and track coverage trends. > > Role breakdown: > | Skill | Timing | Scope | > |-------|--------|-------| > | /coverage (this skill) | Before/after changes, regular queries | Coverage status + uncovered areas + trends | > | /audit | Before deployment | Full audit | > | /impact | Before changes | Impact analysis |

---

Absolute Prohibitions

1. Do not modify code directly — that is the implementation agent's responsibility
2. Do not change test configuration — only perform coverage analysis
3. Do not arbitrarily lower coverage thresholds — follow the criteria defined in planning documents

---

Actions to Take Immediately on Trigger

1. Detect project type (Python/Node.js/Rust/Go, etc.)
2. Check test environment
3. Collect coverage data
4. Identify uncovered files/areas
5. Analyze coverage trends (when available)
6. Generate report and present recommendations

---

Commands

| Command | Description | |---------|-------------| | /coverage | Full project coverage query | | /coverage | Coverage for a specific directory/file | | /coverage --uncovered | Show only uncovered areas | | /coverage --threshold | Show files below n% coverage | | /coverage --trend | Coverage trend (7-day/30-day) | | /coverage --function | Per-function coverage for a file | | /coverage --branch | Detailed branch coverage | | /coverage --report | Generate detailed HTML report |

---

Execution Process

Step 1: Detect Project Type

ls pyproject.toml setup.py requirements.txt 2>/dev/null && echo "Python"
ls package.json 2>/dev/null && echo "Node.js"
ls Cargo.toml 2>/dev/null && echo "Rust"
ls go.mod 2>/dev/null && echo "Go"

Step 2: Collect Coverage Data

| Project | Command | |---------|---------| | Python | pytest --cov=. --cov-report=json --cov-report=term-missing | | Node.js (Vitest) | npm run test -- --coverage | | Node.js (Jest) | npm run test -- --coverage | | Rust | cargo tarpaulin --out Json | | Go | go test -coverprofile=coverage.out ./... |

Step 3: Identify Uncovered Areas

From the coverage report, identify:

• Uncovered lines
• Uncovered branches
• Uncovered functions

---

Recommendation Thresholds

| Coverage | Verdict | Action | |----------|---------|--------| | 80%+ | Good | Maintain this level during maintenance | | 70–80% | Improve | Add exception/edge case tests | | Below 70% | Low | Run /coverage --uncovered to identify gaps, then add tests |

---

Related Skill Integration

| Skill | When to Link | Purpose | |-------|--------------|---------| | /impact | Before a change | Check coverage after impact analysis | | /audit | Before deployment | Coverage validation within full audit | | /checkpoint | On task completion | Coverage check during code review |

---

Reference Documents

For detailed output examples, per-language configuration, and usage scenarios, see:

• references/output-formats.md — Output format details
• references/tech-specs.md — Per-language test runner configuration
• references/scenarios.md — Usage scenarios and integration examples

---

Last Updated: 2026-03-12 (v2.3.0 - Progressive Disclosure applied)

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: insightflo
• Source: insightflo/claude-impl-tools
• License: MIT

Install and usage instructions live in the source repository linked above.