@glrs-dev/assume 0.14.0 → 0.15.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +20 -0
- package/package.json +5 -5
package/CHANGELOG.md
CHANGED
|
@@ -1,5 +1,25 @@
|
|
|
1
1
|
# @glrs-dev/assume
|
|
2
2
|
|
|
3
|
+
## 0.15.1
|
|
4
|
+
|
|
5
|
+
### Patch Changes
|
|
6
|
+
|
|
7
|
+
- [#302](https://github.com/iceglober/glrs/pull/302) [`2ad94cd`](https://github.com/iceglober/glrs/commit/2ad94cd4efab27d6f0c60c18ad3778f02ed15c1a) Thanks [@iceglober](https://github.com/iceglober)! - Soften the stale-session guidance from `gsa run_with_credentials` so an agent can self-recover in interactive sessions.
|
|
8
|
+
|
|
9
|
+
The stale-session message now tells the agent it MAY launch `gsa login <provider>` in the background (which opens the browser for the user to complete) and then poll the `check_session` tool until valid, rather than always asking the user to run the command. In headless/remote contexts it still falls back to asking the user. Retries are gated on `check_session` reporting valid.
|
|
10
|
+
|
|
11
|
+
## 0.15.0
|
|
12
|
+
|
|
13
|
+
### Minor Changes
|
|
14
|
+
|
|
15
|
+
- [#300](https://github.com/iceglober/glrs/pull/300) [`a091e0c`](https://github.com/iceglober/glrs/commit/a091e0c400bc3ef555e25d3114d8eebabd563dc0) Thanks [@iceglober](https://github.com/iceglober)! - `gsa` MCP server now detects stale sessions and fails fast instead of hanging, and adds a `check_session` tool.
|
|
16
|
+
|
|
17
|
+
Previously, when the AWS SSO session expired mid-run, `run_with_credentials` would hang on the dead credential endpoint until the MCP client gave up with an opaque `-32001 Request timed out` — the agent couldn't tell a stale session from a genuine command failure, and had no way to prompt re-auth.
|
|
18
|
+
|
|
19
|
+
- **Stale-session guard:** before running (and again if the command times out or fails), `run_with_credentials` checks session health via the existing needs-login marker / refresh-token expiry. When the session can't produce credentials it returns a structured `{ session_stale: true, action: "gsa login aws", … }` result and fires a desktop notification, instead of hanging or returning a raw error.
|
|
20
|
+
- **Execution timeout:** new optional `timeout_ms` (default 120000, max 600000). A hung command is killed and returns a clear timeout — which also re-checks session health — rather than blocking until the client times out. Accepts a number or numeric string.
|
|
21
|
+
- **`check_session` tool:** returns `{ valid, needs_login, session_expires_at, refresh_expires_at, action }` so an agent can verify the session at start-of-task or after a failure. Re-auth itself remains a user-run browser flow (`gsa login aws`); the agent can now detect and surface the need immediately.
|
|
22
|
+
|
|
3
23
|
## 0.14.0
|
|
4
24
|
|
|
5
25
|
### Minor Changes
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@glrs-dev/assume",
|
|
3
|
-
"version": "0.
|
|
3
|
+
"version": "0.15.1",
|
|
4
4
|
"publishConfig": {
|
|
5
5
|
"access": "public"
|
|
6
6
|
},
|
|
@@ -51,9 +51,9 @@
|
|
|
51
51
|
"typescript": "^5"
|
|
52
52
|
},
|
|
53
53
|
"optionalDependencies": {
|
|
54
|
-
"@glrs-dev/assume-darwin-arm64": "0.
|
|
55
|
-
"@glrs-dev/assume-darwin-x64": "0.
|
|
56
|
-
"@glrs-dev/assume-linux-x64": "0.
|
|
57
|
-
"@glrs-dev/assume-linux-arm64": "0.
|
|
54
|
+
"@glrs-dev/assume-darwin-arm64": "0.15.1",
|
|
55
|
+
"@glrs-dev/assume-darwin-x64": "0.15.1",
|
|
56
|
+
"@glrs-dev/assume-linux-x64": "0.15.1",
|
|
57
|
+
"@glrs-dev/assume-linux-arm64": "0.15.1"
|
|
58
58
|
}
|
|
59
59
|
}
|