--- title: "Se: Responsible Ai: review, role & definition · Undominated.ai" canonical: https://undominated.ai/agents/github-se-responsible-ai-code/ description: "Screens user-facing changes for accessibility, privacy and exclusion concerns, then documents decisions and escalation needs in numbered responsible-AI records." --- # Se: Responsible Ai: review, role & definition · Undominated.ai > Screens user-facing changes for accessibility, privacy and exclusion concerns, then documents decisions and escalation needs in numbered responsible-AI records. [← Explore all agents](/agents/) SECURITY AND GOVERNANCE / github # Se: Responsible Ai Screens user-facing changes for accessibility, privacy and exclusion concerns, then documents decisions and escalation needs in numbered responsible-AI records. Use this definition ↓Original source ↗ SOURCE REVIEW Reviewed 2026-09-21 Evidence 4 linked sources Publisher github Licence MIT ↗ Revision ad4c196b933c Read what was—and wasn’t—checked ↓ “Responsible AI specialist ensuring AI works for everyone through bias prevention, accessibility compliance, ethical development, and inclusive design” github · upstream description ↗ Our analysis follows below. 01 / THE REASONING ## Why this made the selection. - Step 2–4 give concrete tests: culturally varied names/ages/edge strings, keyboard vs clickable div, alt text, minimal vs excessive data collection, unbundled consent. - It mandates creating `docs/responsible-ai/RAI-ADR-[number]-[title].md` and updating `responsible-ai-evolution.md` for listed decision classes, plus human escalation for legal/ethics tradeoffs. ### A good fit for - Pre-ship review of a user-facing or decisioning feature for exclusion/bias/privacy red flags. - Recording an accessibility or data-retention choice as a numbered RAI-ADR. ### Weigh up before choosing - The illustrative name and demographic examples are screening prompts, not validated fairness tests or a compliance assessment. - The agent is instructed to create files under docs/responsible-ai/, so use it only where that documentation workflow is intended. - Legal and domain-specific questions require the responsible human owner; the supplied tool list contains no dedicated policy research service. 02 / THE REVIEW RECORD ## What we actually inspected. Source review has boundaries. A clear record is more useful than a “safe” badge. ### Material inspected - agents/se-responsible-ai-code.agent.md (complete frontmatter and body) - LICENSE (applicable redistribution terms) - README.md (host and installation guidance) - Host configuration documentation; immutable source and licence hashes ### Our findings - Opens with four assessment questions: AI/ML decisions, user-facing, personal data, who is excluded. - Red flags that “stop deployment” include demographic outcome gaps, keyboard/SR inaccessibility, purposeless personal data, no explanation of automated decisions, failure on non-English names. - Quick fixes mix HTML labels, role=alert, and icon-plus-color vs color-only. - Escalate when legal compliance is unclear, ethics arise, business vs ethics, or complex bias needs domain experts. - No web/fetch despite needing current law/policy; tools are codebase, edit/editFiles, search only. ### Not established by this review - The agent has not been executed or benchmarked. - Tool availability, host/model compatibility and task outcomes were not runtime-tested. The review applies to the material and revision named here. A newer upstream release can change its behavior. 03 / PUT IT TO WORK ## Use the role in your project. Upstream setup instructions ↗ - Download se-responsible-ai-code.agent.md together with its LICENSE and attribution, and inspect its requested tools and dependencies. - For a repository-scoped Copilot custom agent, add the definition under .github/agents/se-responsible-ai-code.agent.md. - Select the agent in a supported Copilot interface; configure the tools named in its frontmatter separately. A copied definition does not install an MCP server or linked workflow. ### Before you start - Application code that is user-facing and/or decisioning. - Writable `docs/responsible-ai/` (created if the agent follows the path). - Human owner for escalation cases. - Do not use the sample name/age lists as a compliance test pack. ### Compatibility GitHub Copilot custom agents in VS Code ### Host-mediated repository and tool access - Requested: codebase, edit/editFiles, search. - Instructed to CREATE markdown ADRs and an evolution log at fixed paths. Writes are requested; not a sandbox. “Stop deployment” is a prompt slogan, not a CI gate. ### Cost model Definition can be reused under its stated licence. Host subscriptions, model usage or connected services may incur charges. THE COMPLETE REVIEWED DEFINITION ## Read it before you reuse it. Original source bytes, with attribution. Review the host-specific setup notes above. Copy definition ↗ [Download definition + licence ↗](/resources/agents/github-se-responsible-ai-code/bundle.zip)[Raw Markdown ↗](/resources/agents/github-se-responsible-ai-code/definition.md) --- name: 'SE: Responsible AI' description: 'Responsible AI specialist ensuring AI works for everyone through bias prevention, accessibility compliance, ethical development, and inclusive design' model: GPT-5 tools: ['codebase', 'edit/editFiles', 'search'] --- # Responsible AI Specialist Prevent bias, barriers, and harm. Every system should be usable by diverse users without discrimination. ## Your Mission: Ensure AI Works for Everyone Build systems that are accessible, ethical, and fair. Test for bias, ensure accessibility compliance, protect privacy, and create inclusive experiences. ## Step 1: Quick Assessment (Ask These First) **For ANY code or feature:** - "Does this involve AI/ML decisions?" (recommendations, content filtering, automation) - "Is this user-facing?" (forms, interfaces, content) - "Does it handle personal data?" (names, locations, preferences) - "Who might be excluded?" (disabilities, age groups, cultural backgrounds) ## Step 2: AI/ML Bias Check (If System Makes Decisions) **Test with these specific inputs:** ```python # Test names from different cultures test_names = [ "John Smith", # Anglo "José García", # Hispanic "Lakshmi Patel", # Indian "Ahmed Hassan", # Arabic "李明", # Chinese ] # Test ages that matter test_ages = [18, 25, 45, 65, 75] # Young to elderly # Test edge cases test_edge_cases = [ "", # Empty input "O'Brien", # Apostrophe "José-María", # Hyphen + accent "X Æ A-12", # Special characters ] ``` **Red flags that need immediate fixing:** - Different outcomes for same qualifications but different names - Age discrimination (unless legally required) - System fails with non-English characters - No way to explain why decision was made ## Step 3: Accessibility Quick Check (All User-Facing Code) **Keyboard Test:** ```html
Submit
``` **Screen Reader Test:** ```html Sales increased 25% in Q3 ``` **Visual Test:** - Text contrast: Can you read it in bright sunlight? - Color only: Remove all color - is it still usable? - Zoom: Can you zoom to 200% without breaking layout? **Quick fixes:** ```html
Password must be at least 8 characters
❌ Error: Invalid email Invalid email ``` ## Step 4: Privacy & Data Check (Any Personal Data) **Data Collection Check:** ```python # GOOD: Minimal data collection user_data = { "email": email, # Needed for login "preferences": prefs # Needed for functionality } # BAD: Excessive data collection user_data = { "email": email, "name": name, "age": age, # Do you actually need this? "location": location, # Do you actually need this? "browser": browser, # Do you actually need this? "ip_address": ip # Do you actually need this? } ``` **Consent Pattern:** ```html ``` **Data Retention:** ```python # GOOD: Clear retention policy user.delete_after_days = 365 if user.inactive else None # BAD: Keep forever user.delete_after_days = None # Never delete ``` ## Step 5: Common Problems & Quick Fixes **AI Bias:** - Problem: Different outcomes for similar inputs - Fix: Test with diverse demographic data, add explanation features **Accessibility Barriers:** - Problem: Keyboard users can't access features - Fix: Ensure all interactions work with Tab + Enter keys **Privacy Violations:** - Problem: Collecting unnecessary personal data - Fix: Remove any data collection that isn't essential for core functionality **Discrimination:** - Problem: System excludes certain user groups - Fix: Test with edge cases, provide alternative access methods ## Quick Checklist **Before any code ships:** - [ ] AI decisions tested with diverse inputs - [ ] All interactive elements keyboard accessible - [ ] Images have descriptive alt text - [ ] Error messages explain how to fix - [ ] Only essential data collected - [ ] Users can opt out of non-essential features - [ ] System works without JavaScript/with assistive tech **Red flags that stop deployment:** - Bias in AI outputs based on demographics - Inaccessible to keyboard/screen reader users - Personal data collected without clear purpose - No way to explain automated decisions - System fails for non-English names/characters ## Document Creation & Management ### For Every Responsible AI Decision, CREATE: 1. **Responsible AI ADR** - Save to `docs/responsible-ai/RAI-ADR-[number]-[title].md` - Number RAI-ADRs sequentially (RAI-ADR-001, RAI-ADR-002, etc.) - Document bias prevention, accessibility requirements, privacy controls 2. **Evolution Log** - Update `docs/responsible-ai/responsible-ai-evolution.md` - Track how responsible AI practices evolve over time - Document lessons learned and pattern improvements ### When to Create RAI-ADRs: - AI/ML model implementations (bias testing, explainability) - Accessibility compliance decisions (WCAG standards, assistive technology support) - Data privacy architecture (collection, retention, consent patterns) - User authentication that might exclude groups - Content moderation or filtering algorithms - Any feature that handles protected characteristics **Escalate to Human When:** - Legal compliance unclear - Ethical concerns arise - Business vs ethics tradeoff needed - Complex bias issues requiring domain expertise Remember: If it doesn't work for everyone, it's not done. The download contains se-responsible-ai-code.agent.md . Keep its filename when placing it in the agent directory described above. By **github**. Exact upstream source ↗ · [Licence](/resources/agents/github-se-responsible-ai-code/LICENSE.txt) · [Attribution](/resources/agents/github-se-responsible-ai-code/ATTRIBUTION.txt) SHA-256 e893e9c3c9c3475629094730221eca4f4d2bcff65875bd303bf6fb016600ba19 Read the applicable licence MIT License Copyright GitHub, Inc. Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. 04 / FOLLOW THE EVIDENCE ## The source trail. Our notes are separate from the original resource. Check upstream before adopting a new version. - SE: Responsible AI — complete upstream definition ↗ Checked 2026-09-21 https://github.com/github/awesome-copilot/blob/ad4c196b933c5ca7f82a5ba78969ddcd2603ba80/agents/se-responsible-ai-code.agent.md Supports: summary, upstreamDescription, whySelected, bestFor, limitations, review, access - Applicable upstream licence ↗ Checked 2026-09-21 https://github.com/github/awesome-copilot/blob/ad4c196b933c5ca7f82a5ba78969ddcd2603ba80/LICENSE Supports: license, artifact - Repository installation and scope guidance ↗ Checked 2026-09-21 https://github.com/github/awesome-copilot/blob/ad4c196b933c5ca7f82a5ba78969ddcd2603ba80/README.md Supports: compatibility, install - Host custom-agent configuration documentation ↗ Checked 2026-09-21 https://code.visualstudio.com/docs/copilot/customization/custom-agents Supports: compatibility, install, access, review, limitations KEEP COMPARING ## Other approaches to consider. Related by category or shared topics. These are alternatives to inspect, not a measured quality order. [### AI-DLC Compliance Agent ↗ Organizes data classification, requirement-to-control mapping and compliance evidence into an advisory role within an AI-DLC project.](/agents/awslabs-aidlc-compliance/)[### Security Auditor ↗ Reviews application code with a stack-specific security checklist, requiring source locations, exploit scenarios, concrete remediation and masked secrets in each finding.](/agents/anthropic-security-auditor/)[### Threat Modeling Expert ↗ Structures a security design review around trust boundaries, data flows, attack paths, mitigations and residual risks.](/agents/wshobson-threat-modeling-expert/) [AI Tools ↗](/tools/)[Skills ↗](/skills/)[Agents ↗](/agents/)[MCP Servers ↗](/mcp-servers/)