Desktop app: sessions die minutes after spawn — "GitHub credential registration is no longer available for this session" makes the github-mcp-server catalog stale and fatal

[TwoPatient]

Environment

• GitHub Copilot desktop app 1.1.22 (commit c3daa01), macOS arm64, bundled CLI 1.0.84-5 (server mode, stdio, Rust JSON-RPC engine)
• Single github.com account signed in to the app; gh auth status (bundled gh 2.98.0, using the app-exported GH_TOKEN) reports a valid token and gh issue list/gh repo view succeed with it
• Standalone CLI 1.0.86 (/opt/homebrew/bin/copilot) on the same machine/account is not affected

Summary

In desktop-app sessions the per-session GitHub credential registration that the CLI server's built-in github-mcp-server uses (dynamic header provider) disappears ~3–5 minutes after the CLI process spawns. github-mcp-server initialises fine at warm-up, but the first call to any of its tools (web_search, github-mcp-server-*) then fails with a header-provider error, the MCP tool catalog is marked stale, and every subsequent model turn fails until the session ends with session.end reason: error. Background sub-agents in the same process die with it. The app UI reports "you are no longer logged in to github-mcp-server" and links to the MCP list, but the built-in server is not listed there, so there is nothing the user can re-authenticate. Three full app restarts did not help.

Steps to reproduce

1. Create a new worktree session in the desktop app and wait ~3 minutes.
2. Ask the agent to call web_search (or any github-mcp-server-* tool).
3. Observe the tool error, then repeated session errors on the following turns, then the session terminates.

Actual behaviour

Tool result:

MCP tool catalog changed before tool "github-mcp-server-web_search" could be invoked: Failed to verify cached MCP tools: MCP request failed: Transport send error: unexpected server response: dynamic header provider failed: GitHub credential registration is no longer available for this session

Followed by 5 × session.error:

Execution failed: Failed to refresh MCP tools for server github-mcp-server; the last catalog is stale

then session.end with reason: error — the CLI process exits and the worktree session is unusable.

CLI process log (~/.copilot/logs/process-<spawn>-<pid>.log, trimmed):

18:58:12.987Z [INFO] Starting CLI in server mode (stdio)
18:58:14.824Z [INFO] [rust:rmcp::service] Service initialized as client            <- github-mcp-server OK at warm-up
18:58:44.524Z [INFO] Created session: db372b5f-…
19:01:02.593Z [WARNING] [rust:copilot_runtime::session::mcp::tool_catalog] Failed to list MCP tools for github-mcp-server; continuing with the previously listed tools: MCP request failed: Transport send error: Transport [mcp::client::ToolListRevisionTransport<…StreamableHttpClientWorker<…>>] error: unexpected server response: dynamic header provider failed: GitHub credential registration is no longer available for this session
… (30+ repeats until the session terminates)

Processes that never call github-mcp tools show the same root symptom non-fatally, from ~2 s after spawn, in the remote exporter:

[WARNING] [rust:remote::api_session_exporter] remote session batch flush failed {"error":"GitHub credential authority error: GitHub token provider registration \"<per-process uuid>\" is unavailable", …}

Expected behaviour

1. The credential registration should remain valid for the lifetime of the CLI process, or be re-established on demand when the header provider needs it.
2. A failed header refresh / tools/list should surface as a tool error only — it should not invalidate the catalog for unrelated model turns, and it must not terminate the session.
3. If the built-in server's auth is lost, the UI should offer a way to re-authenticate it (the current "no longer logged in" message points to an MCP list that does not contain the built-in server).

Timeline (UTC, 2026-09-18) — four sessions lost

CLI process spawn	First failure	Δ
18:07:08	18:10:06	~3 min
18:11:56	18:15:36	~3.5 min
18:38:57	18:43:46	~5 min
18:58:12	19:01:02	~3 min

The app's accounts.token_updated_at was rewritten at 18:11:42 — inside the failure window — yet the failures continued in processes spawned after that and after each of three app restarts (18:37:51, 18:38:19, 18:55:46).

Also observed 2026-09-16 08:12 (app 1.1.21).

Possible regression: on 2026-09-08 (app 1.1.15) the identical dynamic header provider failed: GitHub credential registration is no longer available for this session error occurred in a sub-agent, but it was a tool-level failure only — the session carried on. Since 1.1.21 the stale catalog is fatal. This looks like two issues: (a) registration loss (older), and (b) fatal stale-catalog handling (newer).

Ruled out

• Prompts/skills: reproduces with a bare web_search call.
• Token validity: the app-exported token works for gh API calls; a standalone copilot -p "…call web_search…" --allow-tool web_search succeeds in ~27 s from the same machine/account.
• Project ↔ account link: the projects are Azure DevOps repos (projects.github_account_id is NULL by design); this only affects create_issue, and the failures reproduce regardless.
• App state: three app restarts, no change. App logs (github-app.<pid>.log) contain no credential/token-provider entries at the failure times.
• Other MCP servers (Azure DevOps stdio, Context7/Microsoft Learn HTTP) keep working in the same sessions.

Impact: any agent workflow that touches web_search or github-mcp-server-* in a desktop-app session is currently unusable; the standalone CLI is the only workaround.

Logs

Full CLI process logs for all four failed sessions plus the non-fatal exporter variant, and the app logs, are available — happy to share them privately on request.

[JamesFescoAVB]

Independent Windows reproduction confirming this is cross-platform and not limited to worktree/child-agent sessions.

Environment

• Windows
• GitHub Copilot desktop app 1.1.22
• Bundled CLI 1.0.84-5
• Copilot Business account (JamesFescoAVB, )

Observations

• The same exact error occurred 71 times in a 6-hour window across 62 Copilot process logs on 2026-09-18.
• It reproduced in a plain general-chat session with no custom agent or child session: calling github-mcp-server-search_code failed with GitHub credential registration is no longer available for this session; shortly afterward the runtime removed the GitHub/Atlassian/web-search MCP tools from the active catalog.
• Failures often arrive in bursts (for example, 6 identical failures within ~630 ms and another 12 within ~1 second). Heavy parallel agent/skill use may amplify the frequency, but the plain-chat reproduction shows concurrency is not required.
• Full process-tree termination and restart temporarily restores operation, but it failed again within minutes during the same diagnostic session.
• A previously used Headroom proxy was removed days earlier; the issue still reproduces. No persistent user- or machine-level HTTP_PROXY, HTTPS_PROXY, ALL_PROXY, GH_TOKEN, or GITHUB_TOKEN variables are configured. GH_TOKEN appears only in the app-created process environment.
• The persisted %LOCALAPPDATA%\Copilot\mcp-tools GitHub catalog is valid/current and contains the expected six tools. The failure occurs at dynamic credential/header refresh, not JSON catalog parsing.
• App logs separately contain repeated failures using the authoritative account credential for Git operations, which may be relevant to the shared credential lifecycle.

This Windows evidence matches the reported fatal stale-catalog behavior. Heavy agent/skill usage appears to make the impact more frequent, but is not the root prerequisite.

[jzm0720]

100% happening to me the last few days. Seems like a new issue with the latest release. Frustrating for sure!

[JamesFescoAVB]

100% happening to me the last few days. Seems like a new issue with the latest release. Frustrating for sure!

I had to switch over to the CLI version. It's working for me, sees my sessions and my agents / skills, but it's not as easy to navigate as the standalone tool. The standalone requires a restart every 10 minutes or so, unworkable for a large task with many subagents.

[palewire]

Also reproducing this on brand new sessions, not just after several minutes:

• App version: 1.1.22
• OS: macOS 26.7 (build 25G229)

For me the "Execution failed: Failed to refresh MCP tools for server github-mcp-server; the last catalog is stale" error hits on the first tool call in a freshly opened chat — starting a new chat within the same running app instance doesn't reliably avoid it. A full quit-and-relaunch of the app sometimes helps temporarily, but the error comes back quickly, sometimes within the same session's first few tool calls. Given the root cause described here (per-session GitHub credential registration expiring 3-5 min after the CLI process spawns), it seems like in my case either the underlying CLI process is being reused/not actually fresh across "new chats," or the credential window is even shorter than 3-5 minutes for some sessions. Happy to grab logs if useful.