Source & Verification Ledger

Every entry shows who published the source, what kind of evidence it is, when KyenAI last checked it, when editorial review is due, and which guide or article relies on it. A review-due label is a maintenance signal—not a claim that the underlying source is false.

Evidence coverage

One ledger for guides and articles

Product documentation is scheduled for review every 14 days; announcements and independent reporting every 30 days; standards, research, and primary records every 90 days.

Sources: 84
Publishers: 25
Guides covered: 20
Articles covered: 12
Review due: 10
Verification needed: 59
Superseded: 0

How to read the ledger

High confidence means an official publisher, primary record, or standards body owns the source. Medium confidence marks independent reporting or preprint research that still needs direct attribution. “Last checked” uses a source-level verification date only when one is recorded for that exact source-and-page use. Editing a page does not refresh its evidence, and checking a shared source for one page does not verify another page that cites it. “Next review” is KyenAI’s editorial schedule, not the source’s expiration date. A source remains visible if it is later superseded so readers can audit historical claims.

CursorVerification needed

Organizations for Cursor Enterprise

Source type: Official announcement
Confidence: High
Published: Jun 3, 2026
Last checked: Oct 14, 2026
Next review: Nov 13, 2026
Cadence: 30 days
Superseded: No record

Used by

Cursor Enterprise Data Retention & Security: What Admins Must VerifyArticle · checked Oct 14, 2026 · 14 cited passagesCited by Security & Governance coverage.Conclusion: Organizations, teams, and groups provide separate governance, budget, model-access, and agent-permission controls; the most permissive setting can win across memberships.Change note: Rechecked the organization/team/group hierarchy and the most-permissive-setting note against the current Cursor changelog page.
AI Coding Agent Governance Checklist: Permissions, Logs and ApprovalsGuide · verification neededOfficial product reference for organization-level governance controls.

Microsoft Visual Studio CodeVerification needed

Choose and use an agent harness

Source type: Independent reporting
Confidence: Medium
Published: Not recorded
Last checked: Oct 14, 2026
Next review: Nov 13, 2026
Cadence: 30 days
Superseded: No record

Used by

AI Coding Agent Harness Engineering: A Practical Integration GuideGuide · checked Oct 14, 2026Current official guide distinguishes harness, model, permissions, and execution target, and documents Copilot, Claude, Codex, and Cloud targets in VS Code.Conclusion: Harnesses remain distinct in tools, models, permissions, customizations, and review workflows; worktree isolation is not itself a security boundary.Change note: Added to anchor the IDE-specific integration and portability cautions.
AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededCurrent official comparison of harness targets, execution locations, permissions, customizations, and review workflows in VS Code.

OpenAIVerification needed

Harness engineering: leveraging Codex in an agent-first world

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Oct 14, 2026
Next review: Nov 13, 2026
Cadence: 30 days
Superseded: No record

Used by

AI Coding Agent Harness Engineering: A Practical Integration GuideGuide · checked Oct 14, 2026Official engineering case study on repository legibility, tools, feedback loops, mechanical architecture checks, and agent autonomy; company-reported results are not generalized benchmarks.Conclusion: The source explicitly emphasizes repository knowledge, executable checks, and observable UI/log feedback as harness investments.Change note: Added as a primary reference for practical harness design.
Loop Engineering for AI Coding Agents: Stop Rules & VerificationGuide · checked Oct 14, 2026Official engineering account of repository context, legibility, architecture checks, UI/log feedback, and agent workflows.Conclusion: The post frames harness work as designing environments, context, and feedback loops around agents; it is an internal case study, not a universal benchmark.Change note: Added as the primary source for the distinction between a repeating loop and the surrounding harness.
AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial engineering case study on repository context, tool feedback, verification, and increasing agent autonomy.

GitHubVerification needed

GitHub Copilot custom instructions support matrix

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Sep 8, 2026
Next review: Sep 22, 2026
Cadence: 14 days
Superseded: No record

Used by

AGENTS.md vs copilot-instructions.md: Which File Should You Use?Guide · checked Sep 8, 2026Lists instruction-file support by Copilot surface, including surface-specific CLAUDE.md support.Conclusion: The matrix still varies by Copilot surface: CLAUDE.md is supported by cloud agent and CLI, while other surfaces list AGENTS.md or Copilot-specific files.Change note: Rechecked the current surface matrix and tightened the wording to avoid a universal yes-or-no claim.
Does GitHub Copilot Read CLAUDE.md? Support Matrix by SurfaceGuide · checked Sep 8, 2026Official matrix for which Copilot surfaces support which instruction files.Conclusion: GitHub currently lists CLAUDE.md for Copilot cloud agent and Copilot CLI, not as the universal instruction file for every Copilot surface.Change note: Rechecked the current matrix and clarified surface-specific ownership.
AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial instruction support matrix across GitHub, IDE, cloud-agent, code-review, and CLI surfaces.
Codex vs GitHub Copilot: Which Coding Agent Fits Your Team?Guide · verification neededOfficial instruction-file support matrix across GitHub, IDE, cloud agent, review, and CLI surfaces.

OpenAIVerification needed

Use AGENTS.md with Codex

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Sep 8, 2026
Next review: Jul 28, 2026
Cadence: 14 days
Superseded: No record

Used by

AGENTS.md Template for Codex, Node.js, Python & MonoreposGuide · checked Sep 8, 2026Official AGENTS.md behavior and precedence reference.Conclusion: OpenAI documents AGENTS.md as repository instructions with nearest-file precedence for Codex workflows.Change note: Rechecked precedence and scope against the current official guide.
AGENTS.md vs copilot-instructions.md: Which File Should You Use?Guide · checked Jul 14, 2026Documents Codex instruction discovery, precedence, scope, and fallback filenames.
AGENTS.md Examples for Codex: Node.js, Python & MonoreposGuide · verification neededOfficial Codex guidance for AGENTS.md scope, precedence, and repository instructions.
AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial root and nested repository-instruction discovery and precedence guidance.
Codex vs Claude CodeGuide · verification neededOfficial repository guidance for AGENTS.md behavior.
Codex vs GitHub Copilot: Which Coding Agent Fits Your Team?Guide · verification neededOfficial Codex repository instruction discovery and scoped precedence guidance.

AGENTS.mdVerification needed

AGENTS.md open format and supported-agent ecosystem

Source type: Standards or methodology
Confidence: High
Published: Not recorded
Last checked: Jul 19, 2026
Next review: Oct 17, 2026
Cadence: 90 days
Superseded: No record

Used by

AGENTS.md vs copilot-instructions.md: Which File Should You Use?Guide · checked Jul 19, 2026Describes the open format, supported-agent ecosystem, 60,000-plus open-source project usage statement, nested guidance, and AAIF stewardship.
AGENTS.md Examples for Codex: Node.js, Python & MonoreposGuide · verification neededCommunity reference for AGENTS.md examples and cross-tool repository guidance.
AI Coding Config Files Guide: AGENTS.md, CLAUDE.md, Gemini, Copilot, Cursor & Windsurf (2026)Guide · verification neededProject overview for the Markdown convention and its use across coding agents; verify each product's own discovery behavior.

AnthropicVerification needed

Claude Code memory documentation

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Jul 14, 2026
Next review: Jul 28, 2026
Cadence: 14 days
Superseded: No record

Used by

AGENTS.md vs copilot-instructions.md: Which File Should You Use?Guide · checked Jul 14, 2026Documents CLAUDE.md loading, scoped project memory, imports such as @AGENTS.md, and symlink options.
AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial CLAUDE.md memory locations, scope, and project instruction behavior.
Does GitHub Copilot Read CLAUDE.md? Support Matrix by SurfaceGuide · verification neededOfficial CLAUDE.md project memory behavior for Claude Code.

Addy OsmaniVerification needed

Loop Engineering

Source type: Independent reporting
Confidence: Medium
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Loop Engineering for AI Coding Agents: Stop Rules & VerificationGuide · verification neededPractitioner overview of moving from prompting agents to designing autonomous loops.

AiderVerification needed

Aider documentation

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

7 Claude Code Alternatives for Different Coding WorkflowsGuide · verification neededOfficial terminal, Git, repository map, test and lint, model provider, scripting, and configuration documentation.

AnthropicVerification needed

Building effective AI agents

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Loop Engineering for AI Coding Agents: Stop Rules & VerificationGuide · verification neededOfficial guidance on agent workflows, evaluator-optimizer loops, and when to add complexity.

AnthropicVerification needed

Claude Code autonomous workflow update

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Codex vs Claude CodeGuide · verification neededOfficial note on Claude Code surfaces, checkpoints, subagents, hooks, and background work.

AnthropicVerification needed

Claude Code CLI reference

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Codex vs Claude CodeGuide · verification neededOfficial reference for permission flags, headless runs, session continuation, and MCP configuration.

AnthropicVerification needed

Claude Code hooks

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Claude Code Hooks vs MCP: Setup, Examples and SecurityGuide · verification neededOfficial hook behavior reference.

AnthropicVerification needed

Claude Code MCP

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Claude Code Hooks vs MCP: Setup, Examples and SecurityGuide · verification neededOfficial MCP setup reference for Claude Code.

AnthropicVerification needed

Claude Code memory

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

AI Coding Config Files Guide: AGENTS.md, CLAUDE.md, Gemini, Copilot, Cursor & Windsurf (2026)Guide · verification neededOfficial CLAUDE.md project-memory and scope guidance for Claude Code.

AnthropicVerification needed

Claude Code overview

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Claude Code subagents workflow examplesGuide · verification neededPlaces subagents in the broader Claude Code workflow.
Loop Engineering for AI Coding Agents: Stop Rules & VerificationGuide · verification neededPlaces scheduling, hooks, subagents, and headless runs in the Claude Code workflow.

AnthropicVerification needed

Claude Code overview

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial overview of Claude Code surfaces, repository work, integrations, and automation paths.
7 Claude Code Alternatives for Different Coding WorkflowsGuide · verification neededOfficial baseline for Claude Code surfaces, repository workflows, integrations, and administration paths.
Codex vs Claude CodeGuide · verification neededOfficial Claude Code workflow overview.

AnthropicVerification needed

Claude Code subagents

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Claude Code subagents workflow examplesGuide · verification neededOfficial subagent capability reference.

AnthropicVerification needed

Create custom subagents

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial subagent roles, tools, permissions, and delegation guidance.

AnthropicVerification needed

Introducing dynamic workflows in Claude Code

Source type: Official announcement
Confidence: High
Published: May 28, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Claude Code Dynamic Workflows Coordinate Parallel SubagentsArticle · verification needed · 10 cited passagesCited by Agent Workflows coverage.

AxiosVerification needed

SpaceX will buy Cursor for $60 billion

Source type: Independent reporting
Confidence: Medium
Published: Jun 16, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Did SpaceX Buy Cursor? $60B Deal Status and TimelineArticle · verification needed · 13 cited passagesCited by AI Coding Agents coverage.

ClineVerification needed

Cline overview

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

7 Claude Code Alternatives for Different Coding WorkflowsGuide · verification neededOfficial editor, terminal, approval, provider, CLI, Kanban, SDK, and enterprise overview.

CursorVerification needed

Cursor Cloud Agents

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial isolated-VM workflow, supported source-control providers, parallel agents, environments, network controls, MCP, hooks, and billing notes.
7 Claude Code Alternatives for Different Coding WorkflowsGuide · verification neededOfficial remote Cloud Agent workflow and current terminology for the feature formerly called Background Agents.

CursorVerification needed

Cursor for iOS

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial Cloud Agent handoff, supervision, review, and availability across mobile, web, and desktop surfaces.

CursorVerification needed

Cursor rules

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial project, user, and team rules plus AGENTS.md and .cursor/rules guidance.

CursorVerification needed

Cursor rules

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

AI Coding Config Files Guide: AGENTS.md, CLAUDE.md, Gemini, Copilot, Cursor & Windsurf (2026)Guide · verification neededOfficial Cursor project rules, metadata, scopes, and legacy .cursorrules migration context.

CursorVerification needed

Introducing Grok 4.5

Source type: Official announcement
Confidence: High
Published: Jul 8, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Did SpaceX Buy Cursor? $60B Deal Status and TimelineArticle · verification needed · 6 cited passagesCited by AI Coding Agents coverage.

Financial TimesVerification needed

SpaceX leapfrogs Amazon to become world's fifth-most valuable company

Source type: Independent reporting
Confidence: Medium
Published: Jun 16, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Did SpaceX Buy Cursor? $60B Deal Status and TimelineArticle · verification needed · 14 cited passagesCited by AI Coding Agents coverage.

GitHubVerification needed

About GitHub Copilot CLI

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial CLI capabilities, permissions, plan mode, custom agents, skills, hooks, MCP, and instruction guidance.
7 Claude Code Alternatives for Different Coding WorkflowsGuide · verification neededOfficial CLI, plan mode, tools, permissions, custom agents, skills, hooks, MCP, and GitHub workflow reference.
Codex vs GitHub Copilot: Which Coding Agent Fits Your Team?Guide · verification neededOfficial CLI capabilities, permissions, plan mode, custom agents, skills, hooks, MCP, and instruction guidance.

GitHubVerification needed

Allowing GitHub Copilot CLI to work autonomously

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Codex vs GitHub Copilot: Which Coding Agent Fits Your Team?Guide · verification neededOfficial autopilot, permission, continuation-limit, safety, and AI Credit considerations.

GitHubVerification needed

Cloud and local sandboxes for GitHub Copilot now in public preview

Source type: Official announcement
Confidence: High
Published: Jun 2, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

GitHub Copilot Sandboxes Put Agent Execution Behind Policy BoundariesArticle · verification needed · 10 cited passagesCited by Security & Governance coverage.
Local vs Cloud AI Coding Agents: Security, Cost and Speed ComparedGuide · verification neededOfficial changelog for local and cloud sandbox behavior.

GitHubVerification needed

Copilot CLI: Improved UI, rubber duck, prompt scheduling, and voice input

Source type: Official documentation
Confidence: High
Published: Jun 2, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Copilot CLI Adds Rubber Duck Review, Prompt Scheduling, and Voice InputArticle · verification needed · 10 cited passagesCited by IDE & CLI coverage.

GitHubVerification needed

Copilot SDK is now generally available

Source type: Official announcement
Confidence: High
Published: Jun 2, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

GitHub Copilot SDK Reaches General AvailabilityArticle · verification needed · 10 cited passagesCited by AI Coding Agents coverage.

GitHubVerification needed

Customize GitHub Copilot with custom instructions

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

AI Coding Config Files Guide: AGENTS.md, CLAUDE.md, Gemini, Copilot, Cursor & Windsurf (2026)Guide · verification neededOfficial Copilot instruction-file paths and surface-specific support, including repository-wide and path-specific guidance.

GitHubVerification needed

GitHub Copilot coding agent

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

AI Coding Agent Governance Checklist: Permissions, Logs and ApprovalsGuide · verification neededOfficial reference for agentic coding workflows and review paths.
Agent Mode vs Chat Mode: Differences, Risks and When to Use EachGuide · verification neededOfficial GitHub concept reference for coding agents.
Local vs Cloud AI Coding Agents: Security, Cost and Speed ComparedGuide · verification neededOfficial reference for asynchronous coding agent workflows.

GitHubVerification needed

GitHub Copilot in Visual Studio Code, May releases

Source type: Official announcement
Confidence: High
Published: Jun 3, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

GitHub Brings Agent-First Workflows Deeper into VS Code StableArticle · verification needed · 10 cited passagesCited by AI Coding Agents coverage.

GitHubVerification needed

Models and pricing for GitHub Copilot

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Codex vs GitHub Copilot: Which Coding Agent Fits Your Team?Guide · verification neededOfficial current model-token and AI Credit billing reference; values should be rechecked before procurement.

GitHubVerification needed

Plans for GitHub Copilot

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Codex vs GitHub Copilot: Which Coding Agent Fits Your Team?Guide · verification neededOfficial current plan positioning and included Copilot capabilities.

GitHubVerification needed

REST API endpoints for search: Search code

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

AI Coding Agent Instruction File Adoption Report — Q3 2026Guide · verification neededPrimary documentation for the public code-search endpoint, result totals, ordering, and API constraints used by the generator.

GitHubVerification needed

Understanding GitHub Code Search syntax

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

AI Coding Agent Instruction File Adoption Report — Q3 2026Guide · verification neededPrimary documentation for filename, path, and extension query qualifiers used in the four searches.

GoogleVerification needed

An important update: Transitioning Gemini CLI to Antigravity CLI

Source type: Official documentation
Confidence: High
Published: May 19, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Google Moves Gemini CLI Users toward Antigravity CLIArticle · verification needed · 10 cited passagesCited by IDE & CLI coverage.
Antigravity CLI migration from Gemini CLIGuide · verification neededOfficial transition announcement.

GoogleVerification needed

Antigravity CLI migration documentation

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Antigravity CLI migration from Gemini CLIGuide · verification neededOfficial migration reference.

GoogleVerification needed

Gemini CLI

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

7 Claude Code Alternatives for Different Coding WorkflowsGuide · verification neededOfficial open-source repository covering terminal tools, GEMINI.md, MCP, scripting, authentication, quotas, and release channels.

GoogleVerification needed

Gemini CLI repository context

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

AI Coding Config Files Guide: AGENTS.md, CLAUDE.md, Gemini, Copilot, Cursor & Windsurf (2026)Guide · verification neededGemini CLI documentation for GEMINI.md context and instruction behavior.

JetBrainsVerification needed

ACP Agent Registry Is Live: Find and Connect AI Coding Agents in Your JetBrains IDE

Source type: Official announcement
Confidence: High
Published: Jan 28, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

JetBrains ACP Agent Registry Opens IDE Access to Multiple Coding AgentsArticle · verification needed · 10 cited passagesCited by Agent Workflows coverage.

KiloVerification needed

What is loop engineering?

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Loop Engineering for AI Coding Agents: Stop Rules & VerificationGuide · verification neededDefines the plan-search-modify-verify-repair cycle for AI-assisted software work.

MCPVerification needed

Model Context Protocol

Source type: Standards or methodology
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 90 days
Superseded: No record

Used by

Claude Code Hooks vs MCP: Setup, Examples and SecurityGuide · verification neededProtocol reference for MCP servers and clients.

MicrosoftVerification needed

Agent mode is now generally available with MCP support

Source type: Official documentation
Confidence: High
Published: Jun 23, 2025
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Visual Studio Agent Mode Adds General Availability and MCP SupportArticle · verification needed · 10 cited passagesCited by IDE & CLI coverage.

MicrosoftVerification needed

VS Code agent mode

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Agent Mode vs Chat Mode: Differences, Risks and When to Use EachGuide · verification neededOfficial VS Code documentation for agent mode behavior.
Local vs Cloud AI Coding Agents: Security, Cost and Speed ComparedGuide · verification neededOfficial local IDE agent mode reference.

OpenAIVerification needed

Codex CLI — Getting Started

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

Codex vs GitHub Copilot: Which Coding Agent Fits Your Team?Guide · verification neededOfficial Codex CLI overview covering local repository work and approval modes.

OpenAIVerification needed

Codex coding-agent instructions

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

AI Coding Config Files Guide: AGENTS.md, CLAUDE.md, Gemini, Copilot, Cursor & Windsurf (2026)Guide · verification neededOfficial AGENTS.md discovery and repository-instruction behavior for Codex.

OpenAIVerification needed

Codex for every role, tool, and workflow

Source type: Official announcement
Confidence: High
Published: Jun 2, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

OpenAI Expands Codex into Plugins, Sites, and AnnotationsArticle · verification needed · 10 cited passagesCited by AI Coding Agents coverage.

OpenAIVerification needed

Introducing GPT-5.3-Codex

Source type: Official announcement
Confidence: High
Published: Feb 5, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

GPT-5.3-Codex Pushes Coding Agents toward Longer-Running WorkArticle · verification needed · 10 cited passagesCited by AI Coding Agents coverage.

OpenAIVerification needed

Introducing the Codex app

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

AI Coding Agents Comparison: Codex, Claude Code, Cursor, and CopilotGuide · verification neededOfficial Codex app description covering threads, worktrees, skills, automations, sandboxing, and cross-surface work.
7 Claude Code Alternatives for Different Coding WorkflowsGuide · verification neededOfficial Codex app description covering threads, worktrees, skills, automations, sandboxing, and cross-surface work.
Codex vs Claude CodeGuide · verification neededOfficial reference for Codex app surfaces, parallel agent threads, worktrees, sandboxing, and availability.
Codex vs GitHub Copilot: Which Coding Agent Fits Your Team?Guide · verification neededOfficial description of Codex app threads, worktrees, skills, automations, sandboxing, surfaces, and plan access.

OpenCodeVerification needed

OpenCode intro

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 14 days
Superseded: No record

Used by

7 Claude Code Alternatives for Different Coding WorkflowsGuide · verification neededOfficial open-source terminal, desktop, IDE, provider, AGENTS.md, Plan mode, permission, and customization overview.

OWASPVerification needed

OWASP Top 10 for LLM Applications

Source type: Standards or methodology
Confidence: High
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 90 days
Superseded: No record

Used by

AI Coding Agent Governance Checklist: Permissions, Logs and ApprovalsGuide · verification neededSecurity reference for common LLM application risks.

ReutersVerification needed

SpaceX locks in $60 billion Cursor deal to close gap with rivals in AI coding race

Source type: Independent reporting
Confidence: Medium
Published: Jun 16, 2026
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Did SpaceX Buy Cursor? $60B Deal Status and TimelineArticle · verification needed · 22 cited passagesCited by AI Coding Agents coverage.

Tom's GuideVerification needed

Tom's Guide Codex vs Claude Code practical comparison

Source type: Independent reporting
Confidence: Medium
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

Codex vs Claude CodeGuide · verification neededPublic practical app-building comparison used as an external example, not KyenAI benchmark data.

WindsurfVerification needed

Windsurf rules

Source type: Independent reporting
Confidence: Medium
Published: Not recorded
Last checked: Not yet verified
Next review: After first verification
Cadence: 30 days
Superseded: No record

Used by

AI Coding Config Files Guide: AGENTS.md, CLAUDE.md, Gemini, Copilot, Cursor & Windsurf (2026)Guide · verification neededOfficial Windsurf Cascade memories and rules documentation; verify current activation and scope behavior before rollout.

GitHubReview due

Adding custom instructions for GitHub Copilot

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Sep 8, 2026
Next review: Aug 13, 2026
Cadence: 30 days
Superseded: No record

Used by

Does GitHub Copilot Read CLAUDE.md? Support Matrix by SurfaceGuide · checked Sep 8, 2026Official repository-wide Copilot instructions path and setup guidance.Conclusion: .github/copilot-instructions.md remains the repository-wide Copilot instructions path.Change note: Rechecked the repository-wide baseline recommendation.
AGENTS.md vs copilot-instructions.md: Which File Should You Use?Guide · checked Jul 14, 2026Documents repository-wide and path-specific Copilot instruction files.

GitHubReview due

GitHub Copilot custom instructions

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Sep 8, 2026
Next review: Oct 8, 2026
Cadence: 30 days
Superseded: No record

Used by

AGENTS.md Template for Codex, Node.js, Python & MonoreposGuide · checked Sep 8, 2026Official repository instruction reference for Copilot workflows.Conclusion: .github/copilot-instructions.md remains GitHub's repository-wide instructions path for Copilot.Change note: Rechecked the Copilot baseline before keeping the cross-tool recommendation.

AnthropicReview due

Connect Claude Code to tools via MCP

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Jul 19, 2026
Next review: Aug 2, 2026
Cadence: 14 days
Superseded: No record

Used by

MCP Server Connected but No Tools? 8 tools/list ChecksGuide · checked Jul 19, 2026Documents /mcp status, tool count, scope precedence, dynamic updates, reconnection, authentication, and Tool Search.

arXivReview due

Configuration Smells in AGENTS.md Files

Source type: Research paper or preprint
Confidence: Medium
Published: Not recorded
Last checked: Jul 19, 2026
Next review: Oct 17, 2026
Cadence: 90 days
Superseded: No record

Used by

AGENTS.md Template for Codex, Node.js, Python & MonoreposGuide · checked Jul 19, 20262026 repository-mining preprint cataloging six instruction-file smells across a sample of 100 AGENTS.md or CLAUDE.md repositories.
AGENTS.md vs copilot-instructions.md: Which File Should You Use?Guide · checked Jul 19, 2026June 2026 preprint that catalogs six instruction-file configuration smells across 100 popular repositories and reports their observed prevalence.

arXivReview due

Evaluating AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?

Source type: Research paper or preprint
Confidence: Medium
Published: Not recorded
Last checked: Jul 19, 2026
Next review: Oct 17, 2026
Cadence: 90 days
Superseded: No record

Used by

AGENTS.md Template for Codex, Node.js, Python & MonoreposGuide · checked Jul 19, 20262026 preprint comparing no context, model-generated context, and developer-written context across coding-agent tasks; reports higher cost and recommends minimal repository-specific guidance.

arXivReview due

Proof-or-Stop: Don't Trust the Agent, Trust the Evidence

Source type: Research paper or preprint
Confidence: Medium
Published: Not recorded
Last checked: Jul 19, 2026
Next review: Oct 17, 2026
Cadence: 90 days
Superseded: No record

Used by

Loop Engineering for AI Coding Agents: Stop Rules & VerificationGuide · checked Jul 19, 2026July 2026 preprint proposing fresh, source-state-bound evidence gates for lifecycle transitions; proof is operational rather than semantic correctness, and the evaluation states material limits.

arXivReview due

Stop Hand-Holding Your Coding Agent: Engineering the Loops that Replace Step-by-Step Prompting

Source type: Research paper or preprint
Confidence: Medium
Published: Not recorded
Last checked: Jul 19, 2026
Next review: Oct 17, 2026
Cadence: 90 days
Superseded: No record

Used by

Loop Engineering for AI Coding Agents: Stop Rules & VerificationGuide · checked Jul 19, 2026June 2026 preprint defining bounded loop specifications with verification, stopping rules, memory, and named terminal states, based partly on a descriptive 50-loop corpus.

GitHubReview due

MCP server debugging guide

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Jul 19, 2026
Next review: Aug 2, 2026
Cadence: 14 days
Superseded: No record

Used by

MCP Server Connected but No Tools? 8 tools/list ChecksGuide · checked Jul 19, 2026Official debugging guidance for server startup, enabled tools, initialize, tools/list, schemas, timeouts, and stdout framing.

CursorReview due

Cursor CLI instruction-file support

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Jul 14, 2026
Next review: Jul 28, 2026
Cadence: 14 days
Superseded: No record

Used by

AGENTS.md vs copilot-instructions.md: Which File Should You Use?Guide · checked Jul 14, 2026Documents Cursor CLI loading root AGENTS.md and CLAUDE.md alongside .cursor/rules.

CursorReview due

Cursor rules documentation

Source type: Official documentation
Confidence: High
Published: Not recorded
Last checked: Jul 14, 2026
Next review: Jul 28, 2026
Cadence: 14 days
Superseded: No record

Used by

AGENTS.md vs copilot-instructions.md: Which File Should You Use?Guide · checked Jul 14, 2026Documents project rules, root AGENTS.md support, and the legacy/deprecated .cursorrules status.

AnthropicCurrent

Harness design for long-running application development

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Oct 14, 2026
Next review: Nov 13, 2026
Cadence: 30 days
Superseded: No record

Used by

AI Coding Agent Harness Engineering: A Practical Integration GuideGuide · checked Oct 14, 2026Official case study on decomposing coding tasks, structured handoffs, and planner/generator/evaluator roles for long-running application work.Conclusion: Anthropic presents a multi-agent approach as one case study and recommends testing harness assumptions and adding only load-bearing complexity.Change note: Added as an independent vendor perspective; not treated as a universal architecture mandate.

AnthropicCurrent

Scaling Managed Agents: Decoupling the brain from the hands

Source type: Official announcement
Confidence: High
Published: Not recorded
Last checked: Oct 14, 2026
Next review: Nov 13, 2026
Cadence: 30 days
Superseded: No record

Used by

AI Coding Agent Harness Engineering: A Practical Integration GuideGuide · checked Oct 14, 2026Official description separates session history, the harness loop/tool router, and the sandbox execution environment.Conclusion: Supports separating the model/harness/sandbox interfaces and revisiting assumptions as models improve.Change note: Added as a source for the architecture-layer definitions.

CursorCurrent

Data Use & Privacy Overview

Source type: Official documentation
Confidence: High
Published: Jun 9, 2026
Last checked: Oct 14, 2026
Next review: Oct 28, 2026
Cadence: 14 days
Superseded: No record

Used by

Cursor Enterprise Data Retention & Security: What Admins Must VerifyArticle · checked Oct 14, 2026 · 6 cited passagesCited by Security & Governance coverage.Conclusion: With Privacy Mode enabled, Cursor states customer data is not used for training and providers use ZDR agreements, subject to abuse-detection and non-ZDR model exceptions.Change note: Rechecked the September 3, 2026 data-use page and retained abuse-detection and non-ZDR model exceptions.

CursorCurrent

Security

Source type: Official documentation
Confidence: High
Published: Apr 24, 2026
Last checked: Oct 14, 2026
Next review: Oct 28, 2026
Cadence: 14 days
Superseded: No record

Used by

Cursor Enterprise Data Retention & Security: What Admins Must VerifyArticle · checked Oct 14, 2026 · 6 cited passagesCited by Security & Governance coverage.Conclusion: Cursor lists current security attestations, enterprise administration controls, Privacy Mode behavior, and trust-portal evidence as of its August 25 update.Change note: Rechecked Privacy Mode and linked enterprise administration areas; the source still displays its August 25, 2026 update date.

CursorCurrent

What's New in Cursor — Latest Updates & Release Notes

Source type: Official announcement
Confidence: High
Published: Oct 6, 2026
Last checked: Oct 14, 2026
Next review: Nov 13, 2026
Cadence: 30 days
Superseded: No record

Used by

Cursor Enterprise Data Retention & Security: What Admins Must VerifyArticle · checked Oct 14, 2026 · 2 cited passagesCited by Security & Governance coverage.Conclusion: The changelog documents Oct 6 remote control for local agents and Sep 23 Rollouts and Security Review; Enterprise remote control is off by default and admins may enable it.Change note: Added a dated table that separates these current workflow features from retention claims and existing security controls.
AI Coding Agent Harness Engineering: A Practical Integration GuideGuide · checked Oct 14, 2026Official changelog dated October 6 and September 23, 2026 for Enterprise remote control, deployment health monitoring, and PR security review.Conclusion: Enterprise remote control is opt-in by admin; Rollouts and Security Review are described as Teams/Enterprise features. Product behavior may vary by plan and rollout.Change note: Added current UI-to-review and delivery signals as vendor-specific examples, not generic harness guarantees.

HarnessCurrent

Harness Software Delivery Platform

Source type: Independent reporting
Confidence: Medium
Published: Not recorded
Last checked: Oct 14, 2026
Next review: Nov 13, 2026
Cadence: 30 days
Superseded: No record

Used by

AI Coding Agent Harness Engineering: A Practical Integration GuideGuide · checked Oct 14, 2026Vendor product page describing Harness's software-delivery platform; product claims and plan coverage require procurement verification.Conclusion: Confirms the separate Harness product/company context; it is not evidence that generic harness engineering requires this vendor.Change note: Added only to disambiguate the company name from the engineering concept.

TypeSafe AICurrent

Introducing System One Models & Jev

Source type: Independent reporting
Confidence: Medium
Published: Not recorded
Last checked: Oct 14, 2026
Next review: Nov 13, 2026
Cadence: 30 days
Superseded: No record

Used by

AI Coding Agent Harness Engineering: A Practical Integration GuideGuide · checked Oct 14, 2026Vendor announcement describing Jev as a typed decision model; performance, reliability, and pricing claims are vendor claims and must be validated on the team's own workload.Conclusion: The vendor describes structured decision outputs rather than code or prose generation; the guide treats outputs as probabilistic signals, not authorization.Change note: Added to distinguish optional decision-model use from coding-agent generation and deterministic controls.

SECCurrent

Space Exploration Technologies Corp. Form 8-K

Source type: Primary filing or record
Confidence: High
Published: Jun 16, 2026
Last checked: Sep 12, 2026
Next review: Dec 11, 2026
Cadence: 90 days
Superseded: No record

Used by

Did SpaceX Buy Cursor? $60B Deal Status and TimelineArticle · checked Sep 12, 2026 · 13 cited passagesCited by AI Coding Agents coverage.Conclusion: The June filing documents the signed transaction and its original closing conditions; it is superseded for current closing status by the August 14 completion filing.Change note: Kept as historical agreement evidence; added the August 14 completion filing as the current legal status source.

SECCurrent

Space Exploration Technologies Corp. Form 8-K: Completion of Acquisition

Source type: Primary filing or record
Confidence: High
Published: Aug 14, 2026
Last checked: Sep 12, 2026
Next review: Dec 11, 2026
Cadence: 90 days
Superseded: No record

Used by

Did SpaceX Buy Cursor? $60B Deal Status and TimelineArticle · checked Sep 12, 2026 · 10 cited passagesCited by AI Coding Agents coverage.Conclusion: The filing states that the merger became effective on August 14, 2026 and Cursor became a wholly owned subsidiary of SpaceX.Change note: Replaced the pending-closing status with the confirmed completion status.

MCPCurrent

MCP authorization security considerations

Source type: Standards or methodology
Confidence: High
Published: Not recorded
Last checked: Sep 8, 2026
Next review: Dec 7, 2026
Cadence: 90 days
Superseded: No record

Used by

MCP Server Security Checklist: 25 Controls for AI Agents — OAuth, Least PrivilegeGuide · checked Sep 8, 2026Official security considerations for token passthrough, SSRF, session threats, and local server compromise.Conclusion: The draft security considerations explicitly cover token passthrough, audience validation, SSRF, token theft, and secure transport.Change note: Kept the source labeled as draft and avoided presenting KyenAI launch controls as protocol mandates.

MCPCurrent

MCP authorization specification

Source type: Standards or methodology
Confidence: High
Published: Not recorded
Last checked: Sep 8, 2026
Next review: Dec 7, 2026
Cadence: 90 days
Superseded: No record

Used by

MCP Server Security Checklist: 25 Controls for AI Agents — OAuth, Least PrivilegeGuide · checked Sep 8, 2026Official requirements for HTTP authorization, protected resource metadata, tokens, scopes, and audience validation.Conclusion: HTTP authorization requires protected-resource discovery and audience-bound token validation.Change note: Rechecked authorization language and retained transport-specific caveats.

MCPCurrent

MCP debugging guide

Source type: Standards or methodology
Confidence: High
Published: Not recorded
Last checked: Sep 8, 2026
Next review: Dec 7, 2026
Cadence: 90 days
Superseded: No record

Used by

MCP Server Connected but No Tools? 8 tools/list ChecksGuide · checked Sep 8, 2026Official guidance for stdio logging, paths, environment variables, initialization, transport debugging, and client logs.Conclusion: The guide continues to separate configuration, transport, initialization, and protocol failures.Change note: Rechecked and aligned the eight-layer fault tree.

MCPCurrent

MCP Inspector

Source type: Standards or methodology
Confidence: High
Published: Not recorded
Last checked: Sep 8, 2026
Next review: Dec 7, 2026
Cadence: 90 days
Superseded: No record

Used by

MCP Server Connected but No Tools? 8 tools/list ChecksGuide · checked Sep 8, 2026Official client-independent tool for inspecting server connections, capabilities, tools, schemas, calls, and notifications.Conclusion: Inspector remains the independent way to determine whether the server exposes expected tools before debugging a specific client.Change note: Rechecked for the minimal-reproduction workflow.
MCP Server Security Checklist: 25 Controls for AI Agents — OAuth, Least PrivilegeGuide · checked Sep 8, 2026Official interactive workflow for validating resources, prompts, tools, notifications, and protocol exchanges.Conclusion: MCP Inspector remains the client-independent validation path for discovery and invocation behavior.Change note: Rechecked the validation workflow for the September refresh.

MCPCurrent

MCP security best practices

Source type: Standards or methodology
Confidence: High
Published: Not recorded
Last checked: Sep 8, 2026
Next review: Dec 7, 2026
Cadence: 90 days
Superseded: No record

Used by

MCP Server Security Checklist: 25 Controls for AI Agents — OAuth, Least PrivilegeGuide · checked Sep 8, 2026Official guidance for least privilege, token security, SSRF, sessions, and local server compromise.Conclusion: The guidance continues to require explicit trust boundaries and covers token, SSRF, session, and local-server risks.Change note: Rechecked for the September security refresh; KyenAI operational controls remain clearly labeled.

MCPCurrent

MCP tools specification

Source type: Standards or methodology
Confidence: High
Published: Not recorded
Last checked: Sep 8, 2026
Next review: Dec 7, 2026
Cadence: 90 days
Superseded: No record

Used by

MCP Server Connected but No Tools? 8 tools/list ChecksGuide · checked Sep 8, 2026Defines tool discovery, tools/list, tool schemas, and list-change capability behavior.Conclusion: A tools-capable server declares the capability and responds to tools/list; listChanged governs dynamic refresh notification.Change note: Rechecked the current draft tools specification and preserved draft status.