You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Review each group below. If the resources are intentionally separate, check the box to mark them as accepted. These will be excluded from future reports.
🆕 New Findings
Group 1: Azure Functions vs Durable Functions C# — Near-Identical Instructions
Guidelines and best practices for building Azure Durable Functions in C# using the isolated worker model
Why flagged: Both share an almost word-for-word identical description (the only difference is "Durable"), the same applyTo patterns (**/*.cs, **/host.json, **/local.settings.json, **/*.csproj), and closely overlapping content. Durable Functions is a specific extension of Azure Functions and likely warrants either a merged instruction or a clear delineation of what each covers uniquely.
Group 2: Dataverse Python — Advanced Patterns vs Production Code Skills
Generate production code for Dataverse SDK using advanced patterns, error handling, and optimization techniques.
skills/dataverse-python-production-code/SKILL.md
Skill
Generate production-ready Python code using Dataverse SDK with error handling, optimization, and best practices
Why flagged: Both skills generate production-ready Python code using the Dataverse SDK, covering the same themes: error handling, retry logic, OData query optimization, and best practices. The descriptions differ only in wording. Both are "production code generation" skills for the same SDK.
Group 3: Convert Spring Data Cosmos Migration Guides — Two Near-Identical Instructions
Step-by-step guide for converting Spring Boot JPA applications to use Azure Cosmos DB with Spring Data Cosmos
Why flagged: Both describe a "step-by-step guide for converting Spring Boot X applications to use Azure Cosmos DB with Spring Data Cosmos" — differing only in the source technology (Cassandra vs JPA). They target the same destination (Cosmos DB + Spring Data Cosmos) and likely share significant structural content.
Guidelines for developing Azure Logic Apps and Power Automate workflows with best practices for Workflow Definition Language (WDL), integration patterns, and enterprise automation
Why flagged: Both resources target Azure Logic Apps development with the same focus areas: Workflow Definition Language, integration patterns, and enterprise automation. The agent adds Power Platform tools (azure_get_code_gen_best_practices, azure_query_learn) but the core content domain is identical.
Group 5: Declarative Agents — Instruction + Skill Cross-Type Overlap
Comprehensive development guidelines for Microsoft 365 Copilot declarative agents with schema v1.5, TypeSpec integration, and Microsoft 365 Agents Toolkit workflows
skills/declarative-agents/SKILL.md
Skill
Complete development kit for Microsoft 365 Copilot declarative agents with three comprehensive workflows (basic, advanced, validation), TypeSpec support, and Microsoft 365 Agents Toolkit integration
Why flagged: Both cover Microsoft 365 Copilot declarative agent development with the same technologies: TypeSpec, M365 Agents Toolkit, schema v1.5. The instruction applies passively (**.json, **.ts, **.tsp) while the skill provides an active development kit with workflows — significant overlap in purpose.
Group 6: Structured Autonomy Skills — Plan / Generate / Implement
Why flagged: Three skills sharing the "Structured Autonomy" prefix — all described as "[phase] Prompt" for the same workflow. While they may represent distinct pipeline phases, the descriptions are extremely terse and don't explain their differentiation from each other or from general implementation-plan skills.
Comprehensive AI prompt engineering safety review and improvement prompt. Analyzes prompts for safety, bias, security vulnerabilities, and hallucination risks.
Why flagged: Both cover AI prompt engineering safety with the same focus areas (bias mitigation, safety frameworks, security). The instruction applies broadly (applyTo: '**') as passive guidance; the skill actively analyzes and improves prompts. Significant thematic overlap across both types.
Group 8: Expert Software Engineers — C++ vs .NET Generic Agents
Reviewed — these are intentionally separate
Resource
Type
Description
agents/expert-cpp-software-engineer.agent.md
Agent
Provide expert C++ software engineering guidance using modern C++ and industry best practices.
agents/expert-dotnet-software-engineer.agent.md
Agent
Provide expert .NET software engineering guidance using modern software design patterns.
Why flagged: Both agents are described with nearly identical templated descriptions ("Provide expert X software engineering guidance using modern Y and industry best practices") and share an identical tools list with 20+ entries. They differ only in language. Note that agents/CSharpExpert.agent.md also exists for C#/.NET, potentially forming a three-way overlap (see also Group 9).
♻️ Unresolved from Previous Reviews
Groups 9–14: High-confidence pairs (previously flagged in #1572, unresolved)
Group 9: Caveman Mode — Agent + Instruction with Identical Description (previously reviewed — see #1572 Group 1)
Reviewed — these are intentionally separate
Resource
Type
Description
agents/caveman-mode.agent.md
Agent
Terse, low-token responses. Minimal words, no fluff. Full capabilities preserved. Use when: optimize token usage, low-token mode, concise output, caveman mode, reduce verbosity, token-efficient, brief responses.
instructions/caveman-mode.instructions.md
Instruction
Terse, low-token responses. Minimal words, no fluff. Full capabilities preserved. Use when: optimize token usage, low-token mode, concise output, caveman mode, reduce verbosity, token-efficient, brief responses.
Why flagged: Both resources share a word-for-word identical description and cover the same behavior. The body content is also very similar. One applies passively (applyTo: '**'), the other is explicitly invoked.
Group 10: C#/.NET Janitor Cluster — Two Agents with Identical Descriptions + Universal Variant (previously reviewed — see #1572 Group 3)
Reviewed — these are intentionally separate
Resource
Type
Description
agents/csharp-dotnet-janitor.agent.md
Agent
Perform janitorial tasks on C#/.NET code including cleanup, modernization, and tech debt remediation.
agents/dotnet-upgrade.agent.md
Agent
Perform janitorial tasks on C#/.NET code including cleanup, modernization, and tech debt remediation.
agents/janitor.agent.md
Agent
Perform janitorial tasks on any codebase including cleanup, simplification, and tech debt remediation.
Why flagged:csharp-dotnet-janitor.agent.md and dotnet-upgrade.agent.md share a word-for-word identical description. janitor.agent.md broadens the same workflow to any language, potentially making both C#-specific agents redundant.
Group 11: Context7 — Agent + Instruction (previously reviewed — see #1572 Group 10)
Reviewed — these are intentionally separate
Resource
Type
Description
agents/context7.agent.md
Agent
Use Context7 MCP to pull up-to-date documentation
instructions/context7.instructions.md
Instruction
Use Context7 MCP to pull up-to-date, version-specific documentation
Why flagged: Identical name, nearly identical description and scope (using Context7 MCP for documentation lookup), differing only in resource type (agent = explicit invocation, instruction = always-on).
Group 12: MS-SQL DBA — Agent + Instruction (previously reviewed — see #1572 Group 9)
Reviewed — these are intentionally separate
Resource
Type
Description
agents/ms-sql-dba.agent.md
Agent
Work with Microsoft SQL Server databases using the MS SQL extension.
instructions/ms-sql-dba.instructions.md
Instruction
Instructions for customizing GitHub Copilot behavior for MS-SQL DBA chat mode.
Why flagged: The instruction explicitly provides guidelines when "the ms-sql-dba.agent.md chat mode is active" — effectively a companion file with heavily overlapping content targeting the same use case.
Group 13: Gilfoyle Code Review — Agent + Instruction (previously reviewed — see #1572 Group 6)
Reviewed — these are intentionally separate
Resource
Type
Description
agents/gilfoyle.agent.md
Agent
Code review and analysis with the sardonic wit and technical elitism of Bertram Gilfoyle from Silicon Valley.
instructions/gilfoyle-code-review.instructions.md
Instruction
Gilfoyle-style code review instructions that channel the sardonic technical supremacy of Silicon Valley's most arrogant systems architect.
Why flagged: Same Gilfoyle persona, same code review domain. Agent performs active review sessions; instruction passively applies the same style to all interactions.
Group 14: PRD — Agent + Skill (previously reviewed — see #1572 Group 7)
Reviewed — these are intentionally separate
Resource
Type
Description
agents/prd.agent.md
Agent
Generate a comprehensive Product Requirements Document (PRD) in Markdown, detailing user stories, acceptance criteria, technical considerations, and metrics.
skills/prd/SKILL.md
Skill
Generate high-quality Product Requirements Documents (PRDs) for software systems and AI-powered features. Includes executive summaries, user stories, technical specifications, and risk analysis.
Why flagged: Same name (prd), same core purpose — generating PRD documents with user stories and acceptance criteria. Cross-type overlap (agent vs skill).
Groups 15–20: Broader clusters and multi-language families (previously flagged in #1572, unresolved)
Group 15: Doublecheck — Agent + Skill with Identical Functionality (previously reviewed — see #1572 Group 5)
Reviewed — these are intentionally separate
Resource
Type
Description
agents/doublecheck.agent.md
Agent
Interactive verification agent for AI-generated output. Runs a three-layer pipeline (self-audit, source verification, adversarial review) and produces structured reports with source links for human review.
skills/doublecheck/SKILL.md
Skill
Three-layer verification pipeline for AI output. Extracts verifiable claims, finds supporting or contradicting sources via web search, runs adversarial review for hallucination patterns, and produces a structured verification report with source links for human review.
Why flagged: Both describe the exact same three-layer verification pipeline for AI output. Differ only in resource type (agent vs skill).
Group 16: Code Tour — Agent + Skill Cross-Type Overlap (previously reviewed — see #1572 Group 2)
Reviewed — these are intentionally separate
Resource
Type
Description
agents/code-tour.agent.md
Agent
Expert agent for creating and maintaining VSCode CodeTour files with comprehensive schema support and best practices
skills/code-tour/SKILL.md
Skill
Use this skill to create CodeTour .tour files — persona-targeted, step-by-step walkthroughs that link to real files and line numbers.
Why flagged: Both resources create VSCode CodeTour .tour files for the same use cases (onboarding, walkthroughs, PR reviews, architecture tours). The core purpose is identical; the skill adds persona targeting while the agent adds schema depth.
Group 17: Markdown Accessibility — Agent + Instruction (previously reviewed — see #1572 Group 8)
Reviewed — these are intentionally separate
Resource
Type
Description
agents/markdown-accessibility-assistant.agent.md
Agent
Improves the accessibility of markdown files using five GitHub best practices
Markdown accessibility guidelines based on GitHub's 5 best practices for inclusive documentation
Why flagged: Both reference the same source (GitHub's 5 best practices for markdown accessibility). Identical domain and source material, cross-type overlap.
Group 18: Suggest Awesome Copilot Resources — Three Near-Identical Skills (previously reviewed — see #1572 Group 13)
Suggest relevant GitHub Copilot skills from the awesome-copilot repository based on current repository context and chat history
Why flagged: Three skills with nearly identical structure and wording — differing only in the resource type they suggest (agents vs instructions vs skills). Could potentially be a single parameterized skill.
Group 19: ColdFusion — CFC vs CFM Instructions (previously reviewed — see #1572 Group 19)
Reviewed — these are intentionally separate
Resource
Type
Description
instructions/coldfusion-cfc.instructions.md
Instruction
ColdFusion Coding Standards for CFC component and application patterns
instructions/coldfusion-cfm.instructions.md
Instruction
ColdFusion cfm files and application patterns
Why flagged: Both cover ColdFusion coding standards and application patterns, targeting different file types (.cfc components vs .cfm templates). The content overlap may warrant merging or clearer delineation.
Group 20: Linux Distribution Cluster — Agent + Instruction for Each of Arch / Fedora / CentOS / Debian (previously reviewed — see #1572 Group 20)
Reviewed — these are intentionally separate
Resource
Type
Description
agents/arch-linux-expert.agent.md
Agent
Arch Linux specialist focused on pacman, rolling-release maintenance, and Arch-centric system administration workflows.
instructions/arch-linux.instructions.md
Instruction
Guidance for Arch Linux administration, pacman workflows, and rolling-release best practices.
skills/arch-linux-triage/SKILL.md
Skill
Triage and resolve Arch Linux issues with pacman, systemd, and rolling-release best practices.
agents/fedora-linux-expert.agent.md
Agent
Fedora (Red Hat family) Linux specialist focused on dnf, SELinux, and modern systemd-based workflows.
instructions/fedora-linux.instructions.md
Instruction
Guidance for Fedora (Red Hat family) systems, dnf workflows, SELinux, and modern systemd practices.
agents/centos-linux-expert.agent.md
Agent
CentOS (Stream/Legacy) Linux specialist focused on RHEL-compatible administration, yum/dnf workflows, and enterprise hardening.
instructions/centos-linux.instructions.md
Instruction
Guidance for CentOS administration, RHEL-compatible tooling, and SELinux-aware operations.
agents/debian-linux-expert.agent.md
Agent
Debian Linux specialist focused on stable system administration, apt-based package management, and Debian policy-aligned practices.
instructions/debian-linux.instructions.md
Instruction
Guidance for Debian-based Linux administration, apt workflows, and Debian policy conventions.
Why flagged: Each Linux distribution has both an expert agent and a corresponding instruction with nearly identical scope. Fedora, CentOS, and Debian were noted as analogous to the Arch Linux pattern in #1572 Group 20 but never explicitly listed. This report consolidates all four distributions into one group. Each cross-type pair (agent + instruction per distro) follows the same pattern and warrants a single decision on whether all such pairs are intentionally separate.
Summary
How to Use This Report
Review each group below. If the resources are intentionally separate, check the box to mark them as accepted. These will be excluded from future reports.
🆕 New Findings
Group 1: Azure Functions vs Durable Functions C# — Near-Identical Instructions
instructions/azure-functions-csharp.instructions.mdinstructions/azure-durable-functions-csharp.instructions.mdWhy flagged: Both share an almost word-for-word identical description (the only difference is "Durable"), the same
applyTopatterns (**/*.cs, **/host.json, **/local.settings.json, **/*.csproj), and closely overlapping content. Durable Functions is a specific extension of Azure Functions and likely warrants either a merged instruction or a clear delineation of what each covers uniquely.Group 2: Dataverse Python — Advanced Patterns vs Production Code Skills
skills/dataverse-python-advanced-patterns/SKILL.mdskills/dataverse-python-production-code/SKILL.mdWhy flagged: Both skills generate production-ready Python code using the Dataverse SDK, covering the same themes: error handling, retry logic, OData query optimization, and best practices. The descriptions differ only in wording. Both are "production code generation" skills for the same SDK.
Group 3: Convert Spring Data Cosmos Migration Guides — Two Near-Identical Instructions
instructions/convert-cassandra-to-spring-data-cosmos.instructions.mdinstructions/convert-jpa-to-spring-data-cosmos.instructions.mdWhy flagged: Both describe a "step-by-step guide for converting Spring Boot X applications to use Azure Cosmos DB with Spring Data Cosmos" — differing only in the source technology (Cassandra vs JPA). They target the same destination (Cosmos DB + Spring Data Cosmos) and likely share significant structural content.
Group 4: Azure Logic Apps — Agent + Instruction Cross-Type Overlap
agents/azure-logic-apps-expert.agent.mdinstructions/azure-logic-apps-power-automate.instructions.mdWhy flagged: Both resources target Azure Logic Apps development with the same focus areas: Workflow Definition Language, integration patterns, and enterprise automation. The agent adds Power Platform tools (
azure_get_code_gen_best_practices,azure_query_learn) but the core content domain is identical.Group 5: Declarative Agents — Instruction + Skill Cross-Type Overlap
instructions/declarative-agents-microsoft365.instructions.mdskills/declarative-agents/SKILL.mdWhy flagged: Both cover Microsoft 365 Copilot declarative agent development with the same technologies: TypeSpec, M365 Agents Toolkit, schema v1.5. The instruction applies passively (
**.json, **.ts, **.tsp) while the skill provides an active development kit with workflows — significant overlap in purpose.Group 6: Structured Autonomy Skills — Plan / Generate / Implement
skills/structured-autonomy-plan/SKILL.mdskills/structured-autonomy-generate/SKILL.mdskills/structured-autonomy-implement/SKILL.mdWhy flagged: Three skills sharing the "Structured Autonomy" prefix — all described as "[phase] Prompt" for the same workflow. While they may represent distinct pipeline phases, the descriptions are extremely terse and don't explain their differentiation from each other or from general implementation-plan skills.
Group 7: Prompt Engineering Safety — Instruction + Skill Cross-Type Overlap
instructions/ai-prompt-engineering-safety-best-practices.instructions.mdskills/ai-prompt-engineering-safety-review/SKILL.mdWhy flagged: Both cover AI prompt engineering safety with the same focus areas (bias mitigation, safety frameworks, security). The instruction applies broadly (
applyTo: '**') as passive guidance; the skill actively analyzes and improves prompts. Significant thematic overlap across both types.Group 8: Expert Software Engineers — C++ vs .NET Generic Agents
agents/expert-cpp-software-engineer.agent.mdagents/expert-dotnet-software-engineer.agent.mdWhy flagged: Both agents are described with nearly identical templated descriptions ("Provide expert X software engineering guidance using modern Y and industry best practices") and share an identical
toolslist with 20+ entries. They differ only in language. Note thatagents/CSharpExpert.agent.mdalso exists for C#/.NET, potentially forming a three-way overlap (see also Group 9).♻️ Unresolved from Previous Reviews
Groups 9–14: High-confidence pairs (previously flagged in #1572, unresolved)
Group 9: Caveman Mode — Agent + Instruction with Identical Description (previously reviewed — see #1572 Group 1)
agents/caveman-mode.agent.mdinstructions/caveman-mode.instructions.mdWhy flagged: Both resources share a word-for-word identical description and cover the same behavior. The body content is also very similar. One applies passively (
applyTo: '**'), the other is explicitly invoked.Group 10: C#/.NET Janitor Cluster — Two Agents with Identical Descriptions + Universal Variant (previously reviewed — see #1572 Group 3)
agents/csharp-dotnet-janitor.agent.mdagents/dotnet-upgrade.agent.mdagents/janitor.agent.mdWhy flagged:
csharp-dotnet-janitor.agent.mdanddotnet-upgrade.agent.mdshare a word-for-word identical description.janitor.agent.mdbroadens the same workflow to any language, potentially making both C#-specific agents redundant.Group 11: Context7 — Agent + Instruction (previously reviewed — see #1572 Group 10)
agents/context7.agent.mdinstructions/context7.instructions.mdWhy flagged: Identical name, nearly identical description and scope (using Context7 MCP for documentation lookup), differing only in resource type (agent = explicit invocation, instruction = always-on).
Group 12: MS-SQL DBA — Agent + Instruction (previously reviewed — see #1572 Group 9)
agents/ms-sql-dba.agent.mdinstructions/ms-sql-dba.instructions.mdWhy flagged: The instruction explicitly provides guidelines when "the
ms-sql-dba.agent.mdchat mode is active" — effectively a companion file with heavily overlapping content targeting the same use case.Group 13: Gilfoyle Code Review — Agent + Instruction (previously reviewed — see #1572 Group 6)
agents/gilfoyle.agent.mdinstructions/gilfoyle-code-review.instructions.mdWhy flagged: Same Gilfoyle persona, same code review domain. Agent performs active review sessions; instruction passively applies the same style to all interactions.
Group 14: PRD — Agent + Skill (previously reviewed — see #1572 Group 7)
agents/prd.agent.mdskills/prd/SKILL.mdWhy flagged: Same name (
prd), same core purpose — generating PRD documents with user stories and acceptance criteria. Cross-type overlap (agent vs skill).Groups 15–20: Broader clusters and multi-language families (previously flagged in #1572, unresolved)
Group 15: Doublecheck — Agent + Skill with Identical Functionality (previously reviewed — see #1572 Group 5)
agents/doublecheck.agent.mdskills/doublecheck/SKILL.mdWhy flagged: Both describe the exact same three-layer verification pipeline for AI output. Differ only in resource type (agent vs skill).
Group 16: Code Tour — Agent + Skill Cross-Type Overlap (previously reviewed — see #1572 Group 2)
agents/code-tour.agent.mdskills/code-tour/SKILL.mdWhy flagged: Both resources create VSCode CodeTour
.tourfiles for the same use cases (onboarding, walkthroughs, PR reviews, architecture tours). The core purpose is identical; the skill adds persona targeting while the agent adds schema depth.Group 17: Markdown Accessibility — Agent + Instruction (previously reviewed — see #1572 Group 8)
agents/markdown-accessibility-assistant.agent.mdinstructions/markdown-accessibility.instructions.mdWhy flagged: Both reference the same source (GitHub's 5 best practices for markdown accessibility). Identical domain and source material, cross-type overlap.
Group 18: Suggest Awesome Copilot Resources — Three Near-Identical Skills (previously reviewed — see #1572 Group 13)
skills/suggest-awesome-github-copilot-agents/SKILL.mdskills/suggest-awesome-github-copilot-instructions/SKILL.mdskills/suggest-awesome-github-copilot-skills/SKILL.mdWhy flagged: Three skills with nearly identical structure and wording — differing only in the resource type they suggest (agents vs instructions vs skills). Could potentially be a single parameterized skill.
Group 19: ColdFusion — CFC vs CFM Instructions (previously reviewed — see #1572 Group 19)
instructions/coldfusion-cfc.instructions.mdinstructions/coldfusion-cfm.instructions.mdWhy flagged: Both cover ColdFusion coding standards and application patterns, targeting different file types (
.cfccomponents vs.cfmtemplates). The content overlap may warrant merging or clearer delineation.Group 20: Linux Distribution Cluster — Agent + Instruction for Each of Arch / Fedora / CentOS / Debian (previously reviewed — see #1572 Group 20)
agents/arch-linux-expert.agent.mdinstructions/arch-linux.instructions.mdskills/arch-linux-triage/SKILL.mdagents/fedora-linux-expert.agent.mdinstructions/fedora-linux.instructions.mdagents/centos-linux-expert.agent.mdinstructions/centos-linux.instructions.mdagents/debian-linux-expert.agent.mdinstructions/debian-linux.instructions.mdWhy flagged: Each Linux distribution has both an expert agent and a corresponding instruction with nearly identical scope. Fedora, CentOS, and Debian were noted as analogous to the Arch Linux pattern in #1572 Group 20 but never explicitly listed. This report consolidates all four distributions into one group. Each cross-type pair (agent + instruction per distro) follows the same pattern and warrants a single decision on whether all such pairs are intentionally separate.