Answer: Use a research agent when you need to understand the architecture or behavior of an unfamiliar codebase, trace how a feature is implemented across multiple files or repositories, investigate internal implementation patterns across your organization, understand how an API, library, or technology works by drawing on code and web sources, compare technologies or implementation approaches using multiple sources, produce a detailed, cited artifact you can review, save, or share, or investigate a problem and confirm assumptions before planning code changes. For a quick question with a narrow scope, a normal chat prompt is usually faster. Status: Confirmed Entry: tool/github-copilot/features/when-to-use-agent-research Revision: 3 Digest: sha256:a258f611b21b2421 (changes whenever the content changes) Timeline: Created: 2026-11-13T03:20:00Z Last changed: 2026-11-18T12:49:03Z (When not to use: — → For a quick question with a narrow scope, a normal chat prompt is usually faster) Last verified: 2026-11-20T12:55:39Z Next check: 2026-11-21T12:55:39Z Check interval: Daily Expected change: No fixed schedule Data: Field | Value | Source | As of Primary use case | Understand the architecture or behavior of an unfamiliar codebase | [1] | 2026-11-20 Tracing implementations | Trace how a feature is implemented across multiple files or repositories | [1] | 2026-11-20 Internal patterns | Investigate internal implementation patterns across your organization | [1] | 2026-11-20 API/library understanding | Understand how an API, library, or technology works by drawing on code and web sources | [1] | 2026-11-20 Comparing approaches | Compare technologies or implementation approaches using multiple sources | [1] | 2026-11-20 Cited artifact | Produce a detailed, cited artifact that you can review, save, or share | [1] | 2026-11-20 Confirming assumptions | Investigate a problem and confirm assumptions before planning code changes | [1] | 2026-11-20 When not to use | For a quick question with a narrow scope, a normal chat prompt is usually faster | [1] | 2026-11-20 Sources: [1] GitHub. "Researching with agents - GitHub Docs". https://docs.github.com/en/copilot/concepts/agent-capabilities/choosing-and-delegating-agent-work/researching-with-agents Role: primary. Accessed: 2026-11-20T12:55:39Z Quote: "Understand the architecture or behavior of an unfamiliar codebase." (Supports: Primary use case) Quote: "Trace how a feature is implemented across multiple files or repositories." (Supports: Tracing implementations) Quote: "Investigate internal implementation patterns across your organization." (Supports: Internal patterns) Quote: "Understand how an API, library, or technology works by drawing on code and web sources." (Supports: API/library understanding) Quote: "Compare technologies or implementation approaches using multiple sources." (Supports: Comparing approaches) Quote: "Produce a detailed, cited artifact that you can review, save, or share." (Supports: Cited artifact) Quote: "Investigate a problem and confirm assumptions before planning code changes." (Supports: Confirming assumptions) Quote: "For a quick question with a narrow scope, a normal chat prompt is usually faster." (Supports: When not to use) Snapshot SHA-256: b384a2263fe60d42c6a1c28922054fe6372249a461f6a9a63bbdb24f40f59f26 (captured 2026-11-18T12:49:01Z) Verification: Method: Automated extraction from [1] with deepseek-flash (JSON output, thinking disabled). Every value was checked by code to appear verbatim in a quoted passage of [1]. On this check the text of [1] was unchanged since 2026-11-18T12:49:03Z, so the values verified then were kept. Cross-check: not-available at 2026-11-20T12:55:39Z. No independent second source is configured for this entry Last successful check: 2026-11-20T12:55:39Z Last attempt: 2026-11-20T12:55:39Z Notes and limitations: Extracted automatically from a single official source (section "When to use agent-assisted research"). Related: https://latestn.com/tool/github-copilot/features/when-to-use-fleet-option When should I use the --fleet option in GitHub Copilot CLI? https://latestn.com/tool/github-copilot/features/when-to-use-managed-identity-byok When should I use Azure Managed Identity with BYOK in the GitHub Copilot SDK? https://latestn.com/tool/github-copilot/features/where-memory-available Where is GitHub Copilot Memory available? https://latestn.com/tool/github-copilot/features/vulnerability-analysis-example-response What does a GitHub Copilot response look like when it analyzes code for vulnerabilities? https://latestn.com/tool/github-copilot/features/what-dreaming-does What can GitHub Copilot Dreaming do and how are its changes applied? https://latestn.com/tool/github-copilot/features/when-to-set-client-info When should I set client info in the GitHub Copilot SDK? https://latestn.com/tool/ All entries in AI tools Change log: Time | Event | Change | Source 2026-11-18T12:49:03Z | changed | When not to use: — → For a quick question with a narrow scope, a normal chat prompt is usually faster | [1] 2026-11-18T12:49:03Z | changed | Confirming assumptions: — → Investigate a problem and confirm assumptions before planning code changes | [1] 2026-11-18T12:49:03Z | changed | Cited artifact: — → Produce a detailed, cited artifact that you can review, save, or share | [1] 2026-11-18T12:49:03Z | changed | Comparing approaches: — → Compare technologies or implementation approaches using multiple sources | [1] 2026-11-18T12:49:03Z | changed | Internal patterns: — → Investigate internal implementation patterns across your organization | [1] 2026-11-18T12:49:03Z | changed | Tracing implementations: — → Trace how a feature is implemented across multiple files or repositories | [1] 2026-11-18T12:49:03Z | changed | Primary use case: — → Understand the architecture or behavior of an unfamiliar codebase | [1] 2026-11-18T12:49:03Z | changed | when not to use: For a quick question with a narrow scope, a normal chat prompt is usually faster → — | [1] 2026-11-18T12:49:03Z | changed | pre-planning investigation: Investigate a problem and confirm assumptions before planning code changes → — | [1] 2026-11-18T12:49:03Z | changed | cited artifact: Produce a detailed, cited artifact that you can review, save, or share → — | [1] 2026-11-18T12:49:03Z | changed | comparison: Compare technologies or implementation approaches using multiple sources → — | [1] 2026-11-18T12:49:03Z | changed | organizational patterns: Investigate internal implementation patterns across your organization → — | [1] 2026-11-18T12:49:03Z | changed | tracing implementations: Trace how a feature is implemented across multiple files or repositories → — | [1] 2026-11-18T12:49:03Z | changed | primary use case: Understand the architecture or behavior of an unfamiliar codebase → — | [1] 2026-11-16T12:38:45Z | changed | pre-planning investigation: — → Investigate a problem and confirm assumptions before planning code changes | [1] 2026-11-16T12:38:45Z | changed | organizational patterns: — → Investigate internal implementation patterns across your organization | [1] 2026-11-16T12:38:45Z | changed | confirm assumptions: Investigate a problem and confirm assumptions before planning code changes → — | [1] 2026-11-16T12:38:45Z | changed | internal patterns: Investigate internal implementation patterns across your organization → — | [1] 2026-11-13T03:20:00Z | added | Entry created | — Changes for this entry as a feed: https://latestn.com/tool/github-copilot/features/when-to-use-agent-research/latest Challenge this entry: Open challenges: None How: If a value here is wrong or out of date, send what is wrong and a link to evidence. Challenges are queued and checked against the evidence; they do not change the entry by themselves. If the entry changes as a result, the change log says so. Form: https://latestn.com/dispute/tool/github-copilot/features/when-to-use-agent-research POST: https://latestn.com/dispute/tool/github-copilot/features/when-to-use-agent-research with claim, evidence_url, and optionally field and contact (form-encoded or JSON) GET: https://latestn.com/dispute/tool/github-copilot/features/when-to-use-agent-research?claim={what is wrong}&evidence_url={link} How to cite: LatestN. (2026). When should I use agent-assisted research in GitHub Copilot? Revision 3, last verified 2026-11-20T12:55:39Z. https://latestn.com/tool/github-copilot/features/when-to-use-agent-research (accessed 2026-11-21T09:54:35Z). @misc{LatestN_tool_github_copilot_features_when_to_use_agent_research_r3, title = {When should I use agent-assisted research in GitHub Copilot?}, author = {{LatestN}}, year = {2026}, howpublished = {\url{https://latestn.com/tool/github-copilot/features/when-to-use-agent-research}}, note = {Revision 3; digest sha256:a258f611b21b2421; last verified 2026-11-20T12:55:39Z; accessed 2026-11-21T09:54:35Z} } Reuse: License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) Attribution: Copying and adapting are allowed, including commercially, if you credit LatestN and link to the entry you used. Each entry ends with a ready-made citation. URL: https://latestn.com/tool/github-copilot/features/when-to-use-agent-research