morluto/rea

Fork 0

mirror of https://github.com/morluto/rea.git synced 2026-11-05 06:45:22 +00:00

[GH-ISSUE #3446] feat: add MCP activity dashboard for agent operations #1290

New issue

Open

opened 2026-10-16 21:03:48 +00:00 by bsakaguchi · 1 comment

bsakaguchi commented

2026-10-16 21:03:48 +00:00

Owner

Originally created by @sailormk2 on GitHub (Oct 14, 2026).
Original GitHub issue: https://github.com/morluto/rea/issues/3446

Before submitting

I searched the open and closed issues and pull requests for a related proposal.
I understand the contribution scope must be reviewed before integration; open questions are welcome here.
This request does not include credentials, private samples, or sensitive target data.

Area

MCP or agent integration

Current limitation

When an agent uses REA through MCP, the user cannot see which tool is currently running, whether it is still active, how long it has been running, or which Evidence was produced.

The current dashboard reports environment readiness and results from extension-triggered actions, but it does not expose a live activity view for agent-driven MCP operations.

Use cases

Monitor a long-running native or application analysis without opening a terminal.
Distinguish running, completed, failed, cancelled, and unknown operations.
Review recent REA operations and locate their Evidence IDs.

Desired outcome and success criteria

Add a REA-native activity view for agent-driven MCP operations.

The first implementation may be a bounded CLI view such as rea activity --watch, with a future local dashboard considered if the event model supports it.

The view should show:

the current MCP tool or operation
started, completed, failed, cancelled, or unknown state
elapsed time
Evidence ID when available
bounded recent operation history

CLI and MCP should preserve equivalent activity semantics. Credentials and sensitive target data must not be recorded. The feature must preserve Evidence provenance and report incomplete or unknown state explicitly.

Expected interaction with targets

No target interaction (documentation or usability only)

Target and scope details (optional)

This proposal concerns host-side observability of REA CLI/MCP activity. It does not add target execution, injection, mutation, protection recovery, or automatic data upload. No private samples or credentials are included.

Verification available so far

Idea only; I need help defining checks

Verification details (optional)

No implementation or benchmark is available yet.

Possible checks include:

lifecycle events are emitted for MCP operations
running, completed, failed, cancelled, and unknown states remain distinguishable
CLI and MCP report equivalent activity semantics
Evidence IDs are preserved when available
credentials and sensitive target data are not recorded
activity history remains bounded

Alternatives considered

The current rea doctor command, CLI output, MCP results, retained sessions, and Evidence records were considered.

They provide readiness or completed-result information, but they do not provide a user-facing live view of agent-driven MCP activity while an operation is running.

Possible direction

Define a small REA-owned activity event model shared by CLI and MCP.

A first surface could be a bounded terminal view such as rea activity --watch. A local dashboard may be added later if the event model and retention policy support it.

The implementation should remain host-side and observational. It should not add target execution, injection, mutation, protection recovery, or automatic data upload.

References

No response

Originally created by @sailormk2 on GitHub (Oct 14, 2026). Original GitHub issue: https://github.com/morluto/rea/issues/3446 ### Before submitting - [x] I searched the open and closed issues and pull requests for a related proposal. - [x] I understand the [contribution scope](https://github.com/morluto/rea/blob/main/CONTRIBUTING.md#contribution-scope) must be reviewed before integration; open questions are welcome here. - [x] This request does not include credentials, private samples, or sensitive target data. ### Area MCP or agent integration ### Current limitation When an agent uses REA through MCP, the user cannot see which tool is currently running, whether it is still active, how long it has been running, or which Evidence was produced. The current dashboard reports environment readiness and results from extension-triggered actions, but it does not expose a live activity view for agent-driven MCP operations. ### Use cases - Monitor a long-running native or application analysis without opening a terminal. - Distinguish running, completed, failed, cancelled, and unknown operations. - Review recent REA operations and locate their Evidence IDs. ### Desired outcome and success criteria Add a REA-native activity view for agent-driven MCP operations. The first implementation may be a bounded CLI view such as `rea activity --watch`, with a future local dashboard considered if the event model supports it. The view should show: - the current MCP tool or operation - started, completed, failed, cancelled, or unknown state - elapsed time - Evidence ID when available - bounded recent operation history CLI and MCP should preserve equivalent activity semantics. Credentials and sensitive target data must not be recorded. The feature must preserve Evidence provenance and report incomplete or unknown state explicitly. ### Expected interaction with targets No target interaction (documentation or usability only) ### Target and scope details (optional) This proposal concerns host-side observability of REA CLI/MCP activity. It does not add target execution, injection, mutation, protection recovery, or automatic data upload. No private samples or credentials are included. ### Verification available so far Idea only; I need help defining checks ### Verification details (optional) No implementation or benchmark is available yet. Possible checks include: - lifecycle events are emitted for MCP operations - running, completed, failed, cancelled, and unknown states remain distinguishable - CLI and MCP report equivalent activity semantics - Evidence IDs are preserved when available - credentials and sensitive target data are not recorded - activity history remains bounded ### Alternatives considered The current `rea doctor` command, CLI output, MCP results, retained sessions, and Evidence records were considered. They provide readiness or completed-result information, but they do not provide a user-facing live view of agent-driven MCP activity while an operation is running. ### Possible direction Define a small REA-owned activity event model shared by CLI and MCP. A first surface could be a bounded terminal view such as `rea activity --watch`. A local dashboard may be added later if the event model and retention policy support it. The implementation should remain host-side and observational. It should not add target execution, injection, mutation, protection recovery, or automatic data upload. ### References _No response_

bsakaguchi added the

triage

enhancement

labels

2026-10-16 21:03:48 +00:00

bsakaguchi commented

2026-10-16 21:03:49 +00:00

Author

Owner

@lhaejong-alt commented on GitHub (Oct 14, 2026):

Implemented the first host-side activity observability slice on local branch feat/mcp-activity-dashboard (commit f230b8333).

Current scope:

Records MCP tools/call start and terminal status at the shared transport boundary.
Distinguishes running, completed, failed, cancelled, and unknown.
Preserves duration and an Evidence ID when the result exposes one.
Stores only bounded local JSONL activity metadata; caller arguments, credentials, and target content are not recorded.
Adds rea activity for a snapshot and rea activity --watch for a terminal live view.
Adds Korean documentation in docs/mcp-activity.ko.md.

Validation completed:

npm run check:fast
focused ActivityJournal and MCP transport tests: 5 tests passed
npm run docs:check
compiled CLI smoke check: rea activity --help and JSON snapshot

This is intentionally a first implementation for maintainer review; a web dashboard and further retention/client semantics remain open questions.

@lhaejong-alt commented on GitHub (Oct 14, 2026): Implemented the first host-side activity observability slice on local branch `feat/mcp-activity-dashboard` (commit `f230b8333`). Current scope: - Records MCP `tools/call` start and terminal status at the shared transport boundary. - Distinguishes `running`, `completed`, `failed`, `cancelled`, and `unknown`. - Preserves duration and an Evidence ID when the result exposes one. - Stores only bounded local JSONL activity metadata; caller arguments, credentials, and target content are not recorded. - Adds `rea activity` for a snapshot and `rea activity --watch` for a terminal live view. - Adds Korean documentation in `docs/mcp-activity.ko.md`. Validation completed: - `npm run check:fast` - focused ActivityJournal and MCP transport tests: 5 tests passed - `npm run docs:check` - compiled CLI smoke check: `rea activity --help` and JSON snapshot This is intentionally a first implementation for maintainer review; a web dashboard and further retention/client semantics remain open questions.

bsakaguchi referenced this issue

2026-10-17 21:10:11 +00:00

[PR #1303] [MERGED] feat(ghidra): support native PE DLLs on Windows #2783

bsakaguchi referenced this issue

2026-10-17 21:10:33 +00:00

[PR #1342] [CLOSED] feat(ghidra): admit native x86 and x86-64 PE DLLs on Windows #2811

bsakaguchi added

enhancement

and removed

enhancement

labels

2026-10-31 00:30:28 +00:00