API Deprecation Migration Risk Scorer

Turn imported API changelog, release-note, docs-diff, GitHub release, RSS, spreadsheet, or manual rows into deterministic migration-risk briefs.

This Actor is for teams that already collect upstream API or SDK change evidence and need a structured way to triage migration risk: what changed, which APIs might be affected, how urgent it is, who should own it, and what to do next.

What it does

• Normalizes API/developer-doc change rows into stable dataset items.
• Classifies rows as breaking changes, deprecations, security-update reviews, feature additions, docs updates, or watchlist notices.
• Scores migration risk from severity, deadlines, tracked API matches, and critical-system matches.
• Parses common deadline formats such as ISO dates, human dates, and quarters.
• Produces evidence snippets, owner actions, migration checklists, risk flags, and warnings.

Input

The MVP supports items mode only. Each row may come from a changelog monitor, docs diff, GitHub release, RSS export, spreadsheet, or manual research.

Common row fields:

• vendorName, provider, vendor, serviceName, companyName
• source, feedSource, platform, repository, docsSource
• changeType, type, category, kind
• severity, impact, priority, level
• changeTitle, title, summary, releaseTitle
• description, body, message, details, releaseNotes, changelogText, diff
• affectedApis, apis, endpoints, sdk, package, library, components
• replacementApi, replacement, replacementEndpoint, migrationTarget
• deadlineAt, sunsetDate, removalDate, endOfLife, endOfSupport, deprecationDeadline
• changeUrl, url, sourceUrl, link, htmlUrl

Useful Actor options:

• yourStack: product or integration context used in deterministic wording.
• trackedApis: API names, endpoints, SDKs, or workflow terms to prioritize.
• criticalSystems: customer-critical workflows such as checkout, billing, auth, webhooks, messaging, analytics, or AI inference.
• ownerTeams: teams or roles to route owner actions to.
• maxItems: maximum rows to process.
• scrapedAt: reference timestamp for deadline math and output metadata.

Output

Each dataset item includes:

• normalized vendor/source/title/type/severity/source URL;
• impactScore and urgencyLevel;
• triggerTypes such as breaking-change, deprecation-deadline, security-update-review, migration-planning, or watchlist-context;
• matched tracked APIs and critical systems;
• parsed deadline and days until deadline;
• replacement API, migration effort, evidence snippets, migration brief, owner action, and checklist;
• risk flags and warnings.

Example use cases

• Triage Stripe/OpenAI/GitHub/Shopify/Twilio/AWS changelog rows for integration risk.
• Convert developer-doc diffs into sprint-planning tickets.
• Route deadline-driven API migrations to the owning platform team.
• Prepare support/CS context for customer-facing integration changes.
• Use API/docs changes as cautious technical competitive-intelligence signals.

Limits and safety notes

• This MVP does not scrape or monitor API documentation, changelogs, GitHub releases, RSS feeds, or vendor sites.
• It does not call vendor APIs, GitHub, Apify cloud, browsers, proxies, or external services.
• It does not use AI or LLMs.
• It classifies only the imported row evidence. It does not guarantee complete deprecation coverage, vendor truth, exploitability, legal/compliance status, SLA impact, churn, buyer intent, or migration success.
• Security-update rows are flagged for review only. Validate applicability and exploitability against official vendor sources and your own stack.

Local development

npm install
npm test
apify validate-schema

To run locally without Apify storage, save input as input.json and run:

$node src/main.js --input input.json --output output.json