From 8809e9d4d89b00efcda5592ae3eb51bdbbf870fd Mon Sep 17 00:00:00 2001 From: olaservo Date: Mon, 4 Aug 2025 08:13:32 -0700 Subject: [PATCH 1/8] Add draft Inspector Roadmap --- roadmap.md | 39 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 39 insertions(+) create mode 100644 roadmap.md diff --git a/roadmap.md b/roadmap.md new file mode 100644 index 000000000..ded832cdc --- /dev/null +++ b/roadmap.md @@ -0,0 +1,39 @@ +# MCP Inspector Roadmap + +Timeline is TBD - this preliminary draft is to collect categories of improvements and discuss them with the maintainers group. + +Next steps after deciding on what is highest priority would be to set up a simple Project in the repo and reference the exact issues there. + +## CLI Mode + +We have a few PRs and issues open around these: + +- Add configuration parity with UI mode +- Add feature parity with UI mode + +We've also discussed separating this out into a dedicated repo, if that would help with maintainability. + +## UX Improvements (UI Mode) + +- Revisit UI design? +- Add ability to edit/add arbitrary headers +- TODO: Call out any other specific enhancements/bugs that we think are significant + +## Automation/General Process Improvements + +- Automatic dependency updates with Dependabot (SDK, etc.) +- Automation with Claude Code Github Action: + - Initial pass at PR review + - Initial pass at issue triage/labeling + - Opening PRs for simple bugs + - Needs references to MCP spec/docs and example tools for testing +- Revisit how we want to label issues and PRs +- Playwright tests vs ad-hoc testing - what makes sense to expand here as part of CI vs doing on-demand? + +## LLM Integration + +We previously decided not to include this, but people are finding it useful in alternative testing apps. Our previous concerns were around maintainability and whether this really belongs in Inspector. + +Examples: +- Chat functionality to test MCP features +- Real sampling with an LLM From 66176d7636d7d3bba5dc3e8fbda3f86d51b444da Mon Sep 17 00:00:00 2001 From: olaservo Date: Mon, 4 Aug 2025 08:23:21 -0700 Subject: [PATCH 2/8] Add note on voting on features --- roadmap.md | 1 + 1 file changed, 1 insertion(+) diff --git a/roadmap.md b/roadmap.md index ded832cdc..8873e537e 100644 --- a/roadmap.md +++ b/roadmap.md @@ -28,6 +28,7 @@ We've also discussed separating this out into a dedicated repo, if that would he - Opening PRs for simple bugs - Needs references to MCP spec/docs and example tools for testing - Revisit how we want to label issues and PRs +- Should we be actively taking thumbs up/voting into account for feature requests? - Playwright tests vs ad-hoc testing - what makes sense to expand here as part of CI vs doing on-demand? ## LLM Integration From 931ec7dac8b793dcad78e6212a197f51a04fde47 Mon Sep 17 00:00:00 2001 From: olaservo Date: Tue, 5 Aug 2025 05:05:24 -0700 Subject: [PATCH 3/8] Refine MCP Inspector Roadmap by clarifying sections and improving formatting --- roadmap.md | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) diff --git a/roadmap.md b/roadmap.md index 8873e537e..dab24f899 100644 --- a/roadmap.md +++ b/roadmap.md @@ -1,23 +1,22 @@ # MCP Inspector Roadmap -Timeline is TBD - this preliminary draft is to collect categories of improvements and discuss them with the maintainers group. +This preliminary draft is to collect categories of improvements and discuss them with the maintainers group. Next steps after deciding on what is highest priority would be to set up a simple Project in the repo and reference the exact issues there. ## CLI Mode -We have a few PRs and issues open around these: - -- Add configuration parity with UI mode -- Add feature parity with UI mode - -We've also discussed separating this out into a dedicated repo, if that would help with maintainability. +- Configuration parity with UI mode +- Feature parity with UI mode +- Would separating the CLI out into a dedicated repo help with maintainability? ## UX Improvements (UI Mode) -- Revisit UI design? -- Add ability to edit/add arbitrary headers -- TODO: Call out any other specific enhancements/bugs that we think are significant +- Not all configs are saved consistently +- Configs file support is inconsistent between modes +- Ability to edit/add arbitrary headers +- Editing long sets of args (or other long values) in the sidebar text boxes is annoying +- Revisit UI design ## Automation/General Process Improvements @@ -29,7 +28,8 @@ We've also discussed separating this out into a dedicated repo, if that would he - Needs references to MCP spec/docs and example tools for testing - Revisit how we want to label issues and PRs - Should we be actively taking thumbs up/voting into account for feature requests? -- Playwright tests vs ad-hoc testing - what makes sense to expand here as part of CI vs doing on-demand? +- Playwright tests vs ad-hoc testing - what makes sense to expand here as part of CI vs doing testing on-demand? + - And what can Claude be instructed to do with Playwright, etc? ## LLM Integration From 40db6a015511a04b226863f92eb5267d1908aef9 Mon Sep 17 00:00:00 2001 From: olaservo Date: Tue, 5 Aug 2025 05:06:23 -0700 Subject: [PATCH 4/8] Clarify testing references in automation section of MCP Inspector Roadmap --- roadmap.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/roadmap.md b/roadmap.md index dab24f899..bb41181ab 100644 --- a/roadmap.md +++ b/roadmap.md @@ -25,7 +25,7 @@ Next steps after deciding on what is highest priority would be to set up a simpl - Initial pass at PR review - Initial pass at issue triage/labeling - Opening PRs for simple bugs - - Needs references to MCP spec/docs and example tools for testing + - Needs references to MCP spec/docs and example tools for testing (Ola has examples of this in personal repos to build on) - Revisit how we want to label issues and PRs - Should we be actively taking thumbs up/voting into account for feature requests? - Playwright tests vs ad-hoc testing - what makes sense to expand here as part of CI vs doing testing on-demand? From d06179dacb67e495367524567520d4cee2e392bc Mon Sep 17 00:00:00 2001 From: olaservo Date: Tue, 5 Aug 2025 05:08:18 -0700 Subject: [PATCH 5/8] Clarify testing references in automation section of MCP Inspector Roadmap --- roadmap.md | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/roadmap.md b/roadmap.md index bb41181ab..8db4ddc95 100644 --- a/roadmap.md +++ b/roadmap.md @@ -25,7 +25,9 @@ Next steps after deciding on what is highest priority would be to set up a simpl - Initial pass at PR review - Initial pass at issue triage/labeling - Opening PRs for simple bugs - - Needs references to MCP spec/docs and example tools for testing (Ola has examples of this in personal repos to build on) + - Needs references to MCP spec/docs and example tools for testing. Examples that could be incorporated into a 'Maintainer Toolkit MCP Server' or something similar: + - https://github.com/olaservo/mcp-advisor + - https://github.com/olaservo/mcp-misc/tree/main/example-tools - Revisit how we want to label issues and PRs - Should we be actively taking thumbs up/voting into account for feature requests? - Playwright tests vs ad-hoc testing - what makes sense to expand here as part of CI vs doing testing on-demand? From b283c9ae8d20080a37febd76c9135da6004a18b2 Mon Sep 17 00:00:00 2001 From: olaservo Date: Tue, 5 Aug 2025 19:39:29 -0700 Subject: [PATCH 6/8] Adding clarifications and improving the server management UI section --- roadmap.md | 3 +++ 1 file changed, 3 insertions(+) diff --git a/roadmap.md b/roadmap.md index 8db4ddc95..cd3a1658e 100644 --- a/roadmap.md +++ b/roadmap.md @@ -4,6 +4,8 @@ This preliminary draft is to collect categories of improvements and discuss them Next steps after deciding on what is highest priority would be to set up a simple Project in the repo and reference the exact issues there. +(Note that these aren't in any priority order yet, just collecting everything in one place.) + ## CLI Mode - Configuration parity with UI mode @@ -17,6 +19,7 @@ Next steps after deciding on what is highest priority would be to set up a simpl - Ability to edit/add arbitrary headers - Editing long sets of args (or other long values) in the sidebar text boxes is annoying - Revisit UI design +- Server selection and management UI - Surface config file in a chooser to allow closing one server and starting another from the UI ## Automation/General Process Improvements From 9684c2b9bd08edf3798a3414c2b2f4e1de43647d Mon Sep 17 00:00:00 2001 From: olaservo Date: Sun, 10 Aug 2025 16:06:56 -0700 Subject: [PATCH 7/8] Fix formatting --- roadmap.md | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/roadmap.md b/roadmap.md index cd3a1658e..05a0d0223 100644 --- a/roadmap.md +++ b/roadmap.md @@ -25,21 +25,22 @@ Next steps after deciding on what is highest priority would be to set up a simpl - Automatic dependency updates with Dependabot (SDK, etc.) - Automation with Claude Code Github Action: - - Initial pass at PR review - - Initial pass at issue triage/labeling - - Opening PRs for simple bugs - - Needs references to MCP spec/docs and example tools for testing. Examples that could be incorporated into a 'Maintainer Toolkit MCP Server' or something similar: - - https://github.com/olaservo/mcp-advisor - - https://github.com/olaservo/mcp-misc/tree/main/example-tools + - Initial pass at PR review + - Initial pass at issue triage/labeling + - Opening PRs for simple bugs + - Needs references to MCP spec/docs and example tools for testing. Examples that could be incorporated into a 'Maintainer Toolkit MCP Server' or something similar: + - https://github.com/olaservo/mcp-advisor + - https://github.com/olaservo/mcp-misc/tree/main/example-tools - Revisit how we want to label issues and PRs - Should we be actively taking thumbs up/voting into account for feature requests? - Playwright tests vs ad-hoc testing - what makes sense to expand here as part of CI vs doing testing on-demand? - - And what can Claude be instructed to do with Playwright, etc? + - And what can Claude be instructed to do with Playwright, etc? ## LLM Integration We previously decided not to include this, but people are finding it useful in alternative testing apps. Our previous concerns were around maintainability and whether this really belongs in Inspector. Examples: + - Chat functionality to test MCP features - Real sampling with an LLM From 32f4dc02a849be40d0407596c12ecc410948fc0f Mon Sep 17 00:00:00 2001 From: Ola Hungerford Date: Mon, 11 Aug 2025 07:43:49 -0700 Subject: [PATCH 8/8] Update roadmap.md --- roadmap.md | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/roadmap.md b/roadmap.md index 05a0d0223..207a8ac93 100644 --- a/roadmap.md +++ b/roadmap.md @@ -21,6 +21,11 @@ Next steps after deciding on what is highest priority would be to set up a simpl - Revisit UI design - Server selection and management UI - Surface config file in a chooser to allow closing one server and starting another from the UI +## OAuth Debugging and Compatibility Checking + +- Outh testing and debugging hiccups: see issues labeled `auth` for examples +- Client auth compatibility checker (See: https://github.com/modelcontextprotocol/inspector/pull/694/files) + ## Automation/General Process Improvements - Automatic dependency updates with Dependabot (SDK, etc.)