[PR #1452] [MERGED] docs(mcp): record real Windows first-query recovery #2896

New issue

Closed

opened 2026-10-17 21:11:32 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-17 21:11:32 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/morluto/rea/pull/1452
Author: @N0zoM1z0
Created: 10/10/2026
Status: ✅ Merged
Merged: 10/10/2026
Merged by: @N0zoM1z0

Base: main ← Head: docs/windows-first-query-recovery-1065

📝 Commits (1)

fb0fe9e docs(mcp): record real Windows first-query recovery

📊 Changes

1 file changed (+21 additions, -4 deletions)

View changed files

📝 docs/mcp-contracts.md (+21 -4)

📄 Description

The Ghidra first-query recovery guide previously recorded only a Linux run. Add the real Windows x64 stdio MCP acceptance: a default 60-second client request times out; closing and reopening the same target/provider on the same connection allows a caller-selected 240-second request to complete. Record the cold/warm observations and the lack of intermediate import progress so callers can choose their request deadline accurately.

Verification used a frozen bb6fa1c6 runtime/source snapshot, freshly built 6.3.0 Windows native controls, Node 22.19.0, Ghidra 12.1.4 and JDK 21, all in an owned D:\tmp\rea test directory:

Source-owned 1,024-byte native PE; input SHA-256 unchanged.
Default request: REQUEST_TIMEOUT after 60,015 ms; close: 646 ms.
Reopened cold overview: 163,025 ms; warm overview: 18.7 ms; both report two procedures/two segments.
Start marker before timeout, start/completion markers on successful retry, no intermediate import/auto-analysis progress.
Final close and ping pass; no owned runtime root or observed process-family member remains. Guard exit 0, sampled family peak 754,233,344 bytes.

This documents one small-target recovery flow. It does not establish mutation authority, macOS recovery, broad Windows analysis coverage, or a universal 240-second allowance. Real Hopper verification was not performed. Real Ghidra verification was performed on Windows as described.

Refs #1065 and #1048.

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/morluto/rea/pull/1452 **Author:** [@N0zoM1z0](https://github.com/N0zoM1z0) **Created:** 10/10/2026 **Status:** ✅ Merged **Merged:** 10/10/2026 **Merged by:** [@N0zoM1z0](https://github.com/N0zoM1z0) **Base:** `main` ← **Head:** `docs/windows-first-query-recovery-1065` --- ### 📝 Commits (1) - [`fb0fe9e`](https://github.com/morluto/rea/commit/fb0fe9e20e2fb26046e1bb36314946afd2b0aa2d) docs(mcp): record real Windows first-query recovery ### 📊 Changes **1 file changed** (+21 additions, -4 deletions) <details> <summary>View changed files</summary> 📝 `docs/mcp-contracts.md` (+21 -4) </details> ### 📄 Description The Ghidra first-query recovery guide previously recorded only a Linux run. Add the real Windows x64 stdio MCP acceptance: a default 60-second client request times out; closing and reopening the same target/provider on the same connection allows a caller-selected 240-second request to complete. Record the cold/warm observations and the lack of intermediate import progress so callers can choose their request deadline accurately. Verification used a frozen `bb6fa1c6` runtime/source snapshot, freshly built 6.3.0 Windows native controls, Node 22.19.0, Ghidra 12.1.4 and JDK 21, all in an owned `D:\tmp\rea` test directory: - Source-owned 1,024-byte native PE; input SHA-256 unchanged. - Default request: `REQUEST_TIMEOUT` after 60,015 ms; close: 646 ms. - Reopened cold overview: 163,025 ms; warm overview: 18.7 ms; both report two procedures/two segments. - Start marker before timeout, start/completion markers on successful retry, no intermediate import/auto-analysis progress. - Final close and ping pass; no owned runtime root or observed process-family member remains. Guard exit 0, sampled family peak 754,233,344 bytes. This documents one small-target recovery flow. It does not establish mutation authority, macOS recovery, broad Windows analysis coverage, or a universal 240-second allowance. Real Hopper verification was not performed. Real Ghidra verification was performed on Windows as described. Refs #1065 and #1048. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

bsakaguchi

2026-10-17 21:11:32 +00:00

closed this issue
added the
pull-request
label

bsakaguchi referenced this issue

2026-10-31 01:06:58 +00:00

[PR #2896] [CLOSED] fix(reference-source): classify configured multi-segment generated paths #5478

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

morluto/rea#2896

No description provided.

Rows
Columns