Implements a comprehensive suite of AI SDK v6 tools across multiple categories: - Research (5): page-brief, compare-pages, source-credibility, claim-checklist, timeline-from-text - Web (10): fetch-text, links-catalog, extract-meta, extract-json-ld, redirect-trace, sitemap-read, rss-read, table-extract, robots-policy, url-normalize - Data (15): csv-parse, csv-stringify, json-repair, json-schema-validate, yaml-parse, yaml-stringify, text-chunk, normalize-whitespace, dedupe-by-key, pivot, rows-filter, rows-sort, rows-group-aggregate, rows-join, schema-infer - Doc (12): toc-generate, glossary-build, faq-from-text, executive-brief, decision-record-adr, prd-outline, acceptance-criteria, style-rewrite - Eng (12): diff-text-unified, env-var-docs-generate, dependency-audit-lite, conventional-commit-suggest, markdown-lint-basic, test-case-generate, stacktrace-parse, release-notes, changelog-entry, release-checklist - Security (7): redact-secrets, secret-scan-text, url-risk-heuristic, csp-compose, hardening-checklist-web, access-control-matrix, data-classification-heuristic - Stats (9): effect-size-suite, bootstrap-ci, permutation-test, multiple-testing-adjust, linear-regression-ols, logistic-regression, time-series-decompose-lite, anomaly-detect-mad - Ops (7): slo-draft, runbook-draft, postmortem-draft, postmortem-action-extractor, error-log-triage, coverage-tracker, monitoring-gap-analysis - Agent (15): prompt-to-workflow-skeleton, workflow-validate-io, workflow-explain, workflow-cost-estimate, tool-call-accuracy-score, eval-fixture-build, guardrail-policy-draft, workflow-auto-repair, tool-selection-plan, novelty-score-workflow, workflow-variant-generate, config-normalize, recipe-* - Utility (8): base64-encode, base64-decode, hash-text, regex-extract, template-render, date-parse, json-path-query, url-parse - HTML (3): html-sanitize, html-to-markdown, markdown-to-html All tools follow AI SDK v6 pattern with tool() and jsonSchema<T>(). 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
5.6 KiB
5.6 KiB
Coverage Tracker
Tracks which tools have been used in a workflow and calculates coverage percentage. Useful for testing workflow completeness and analyzing tool utilization patterns.
Installation
npm install @tpmjs/tools-coverage-tracker
Usage
import { coverageTrackerTool } from '@tpmjs/tools-coverage-tracker';
import { generateText } from 'ai';
const result = await generateText({
model: yourModel,
tools: {
trackCoverage: coverageTrackerTool,
},
prompt: 'Track tool usage coverage...',
});
Direct Usage
import { coverageTrackerTool } from '@tpmjs/tools-coverage-tracker';
const result = await coverageTrackerTool.execute({
availableTools: ['searchWeb', 'fetchUrl', 'summarize', 'translateText', 'saveNote'],
usedTools: ['searchWeb', 'fetchUrl', 'searchWeb', 'summarize'],
});
console.log(result);
// {
// coverage: 0.6,
// usedCount: 3,
// totalCount: 5,
// unusedTools: ['translateText', 'saveNote'],
// usedTools: [
// { name: 'searchWeb', used: true, usageCount: 2 },
// { name: 'fetchUrl', used: true, usageCount: 1 },
// { name: 'summarize', used: true, usageCount: 1 }
// ],
// coveragePercent: '60.0%',
// summary: 'Coverage: 60.0% (3/5 tools) | Unused: translateText, saveNote'
// }
Input Schema
{
availableTools: string[]; // All available tool names
usedTools: string[]; // Tools that were actually used (can include duplicates)
}
Output Schema
{
coverage: number; // Coverage ratio (0-1)
usedCount: number; // Number of unique tools used
totalCount: number; // Total number of available tools
unusedTools: string[]; // List of tools not used
usedTools: Array<{ // List of used tools with stats
name: string; // Tool name
used: boolean; // Always true for this array
usageCount: number; // How many times it was called
}>;
coveragePercent: string; // Formatted percentage (e.g., "75.0%")
summary: string; // Human-readable summary
}
Coverage Calculation
Coverage = (Number of unique tools used) / (Total available tools)
- Duplicate tool calls are counted separately in
usageCount - Coverage is based on unique tools used
- Tools used but not in
availableToolstrigger a warning
Use Cases
1. Workflow Testing
Ensure your agent workflow exercises all available tools:
const availableTools = ['search', 'analyze', 'summarize', 'report'];
const usedTools = ['search', 'analyze', 'report'];
const coverage = await coverageTrackerTool.execute({
availableTools,
usedTools,
});
if (coverage.coverage < 0.75) {
console.warn('Low tool coverage - workflow may be incomplete');
}
2. Tool Utilization Analysis
Find which tools are being used most/least:
const result = await coverageTrackerTool.execute({
availableTools: ['tool1', 'tool2', 'tool3', 'tool4'],
usedTools: ['tool1', 'tool1', 'tool1', 'tool2', 'tool2', 'tool3'],
});
// result.usedTools sorted by usage:
// [
// { name: 'tool1', usageCount: 3 },
// { name: 'tool2', usageCount: 2 },
// { name: 'tool3', usageCount: 1 },
// ]
// result.unusedTools: ['tool4']
3. Integration Testing
Track tool coverage across test runs:
const testRuns = [
{ name: 'Test 1', used: ['searchWeb', 'fetchUrl'] },
{ name: 'Test 2', used: ['searchWeb', 'summarize'] },
{ name: 'Test 3', used: ['translateText'] },
];
const allTools = ['searchWeb', 'fetchUrl', 'summarize', 'translateText'];
for (const run of testRuns) {
const coverage = await coverageTrackerTool.execute({
availableTools: allTools,
usedTools: run.used,
});
console.log(`${run.name}: ${coverage.coveragePercent}`);
}
4. Finding Dead Code
Identify tools that are never used:
const coverage = await coverageTrackerTool.execute({
availableTools: ['common', 'rare', 'legacy', 'deprecated'],
usedTools: ['common', 'common', 'common', 'rare'],
});
// coverage.unusedTools: ['legacy', 'deprecated']
// Consider removing these tools
Advanced Features
Duplicate Tracking
The tool tracks how many times each tool is called:
const result = await coverageTrackerTool.execute({
availableTools: ['api1', 'api2'],
usedTools: ['api1', 'api1', 'api1', 'api2'],
});
// result.usedTools[0] = { name: 'api1', usageCount: 3 }
// result.usedTools[1] = { name: 'api2', usageCount: 1 }
Unknown Tool Detection
Warns if tools are used that aren't in the available list:
const result = await coverageTrackerTool.execute({
availableTools: ['tool1', 'tool2'],
usedTools: ['tool1', 'unknownTool'],
});
// result.summary: "... | Warning: 1 unknown tool(s) used"
Examples
Example 1: Full Coverage
const result = await coverageTrackerTool.execute({
availableTools: ['a', 'b', 'c'],
usedTools: ['a', 'b', 'c'],
});
// coverage.coverage = 1.0
// coverage.coveragePercent = "100.0%"
// coverage.unusedTools = []
Example 2: Partial Coverage
const result = await coverageTrackerTool.execute({
availableTools: ['search', 'fetch', 'save', 'analyze'],
usedTools: ['search', 'fetch'],
});
// coverage.coverage = 0.5
// coverage.usedCount = 2
// coverage.unusedTools = ['save', 'analyze']
Example 3: No Coverage
const result = await coverageTrackerTool.execute({
availableTools: ['tool1', 'tool2'],
usedTools: [],
});
// coverage.coverage = 0
// coverage.unusedTools = ['tool1', 'tool2']
License
MIT