Version updated for https://github.com/Tahiram32/downstream-breakage-radar to version v0.5.0. This action is used across all versions by ? repositories. Action Type This is a Composite action. Go to the GitHub Marketplace to find the latest changes. What’s Changed The Enterprise Upgrade (v0.5.0) 🚀 This major update adds powerful compliance, customization, and deprecation lifecycle controls for developers and organizations. 🌟 New Features ⚠️ Deprecation Checker — When a public function or class is removed, the tool checks the base branch to see if it was marked with a deprecation warning first. Pre-deprecated removals are downgraded to medium severity; surprise removals stay high. 🛡️ SARIF Output (--format sarif) — Export results in standardized SARIF JSON format for integration with GitHub’s native Security / Code Scanning dashboard. 📝 API Changelog Generator (--changelog) — Generates a clean breakage-radar-changelog.md listing every public addition, removal, and signature change across Python, Go, and JS/TS files. ⚙️ In-Manifest Configuration — Configure ignored paths, public directories, and severity overrides directly inside pyproject.toml or a breakage-radar.json file — no CLI flags required. 📦 Install / Upgrade pip install --upgrade downstream-breakage-radar 🔧 Usage - name: Scan for breaking changes uses: Tahiram32/downstream-breakage-radar@v0.5.0 with: base-ref: origin/main format: markdown fail-on: high changelog: true 📄 Docs Full documentation and configuration reference: https://tahiram32.github.io/downstream-breakage-radar/

Version updated for https://github.com/Tahiram32/llm-prompt-radar to version v0.2.0. This action is used across all versions by ? repositories. Action Type This is a Composite action. Go to the GitHub Marketplace to find the latest changes. What’s Changed What’s New in v0.2.0 YAML/TOML config file support (.promptradar.yml) LangChain and LlamaIndex SDK support Custom rule definitions (bring-your-own regex) PR comment posting with risk summary table Pre-commit hook integration VS Code extension skeleton pip install --upgrade llm-prompt-radar PyPI: https://pypi.org/project/llm-prompt-radar/0.2.0/

Version updated for https://github.com/Tahiram32/llm-prompt-radar to version v0.1.0. This action is used across all versions by ? repositories. Action Type This is a Composite action. Go to the GitHub Marketplace to find the latest changes. What’s Changed The Missing CI Check for AI-Powered Applications 🛡️ llm-prompt-radar is the first dedicated CI tool for detecting risky changes to LLM prompts and AI configuration before they ship to production. 🌟 What it detects 🚨 Safety guardrail removal — Critical alert when safety/refusal instructions are removed from prompts 🤖 Model downgrades — Flags silent swaps like gpt-4o → gpt-3.5-turbo or claude-3-opus → claude-3-haiku 📝 Prompt file changes — Deep analysis of .prompt, .jinja, .j2 template files with similarity scoring 🔍 In-code system messages — Detects changes to system prompts inside Python, JS, and TS source files ⚙️ LLM parameter changes — Tracks temperature, max_tokens, top_p, and more 🎯 Risk scoring — none / low / medium / high / critical 📊 Multiple output formats — text, json, markdown, github annotations, SARIF 📦 Install pip install llm-prompt-radar 🔧 GitHub Action - uses: actions/checkout@v7 with: fetch-depth: 0 - uses: Tahiram32/llm-prompt-radar@v0.1.0 with: base-ref: origin/main format: github fail-on: high 🔗 Links PyPI: https://pypi.org/project/llm-prompt-radar/0.1.0/ Docs: https://github.com/Tahiram32/llm-prompt-radar#readme