Skip to content

First cut at support for new PAM USS config entries for GitHub. - #2259

Closed
mfordkeeper wants to merge 27 commits into
releasefrom
PR-244-add-github-pam-uss-configs
Closed

First cut at support for new PAM USS config entries for GitHub.#2259
mfordkeeper wants to merge 27 commits into
releasefrom
PR-244-add-github-pam-uss-configs

Conversation

@mfordkeeper

Copy link
Copy Markdown
Contributor

No description provided.

@mfordkeeper
mfordkeeper requested a review from m-afanasiuk August 5, 2026 00:00
@mfordkeeper mfordkeeper self-assigned this Aug 5, 2026
@mfordkeeper
mfordkeeper marked this pull request as draft August 5, 2026 00:00
aaunario-keeper and others added 19 commits August 13, 2026 15:17
…/role policy gates (#2261) (#2263)

* Implement sailpoint changes for transfer user instead and ban eu --delete

* Fix bugcrowd issues
* Add gchat-app-setup for Google Chat Service Mode integration.

Collect and store Google Chat/Pub/Sub credentials in a vault record and generate docker-compose, matching the Slack/Teams setup flow.

* Require a file path for Google Chat service account JSON.

Drop unreliable inline JSON paste at the gchat-app-setup prompt; service account keys are too large for single-line terminal input.

* updated to latest annotations

* addressed review comments

* Validate GChat project ID against Pub/Sub paths and the service account.

Reject full resource names whose project differs from GOOGLE_PROJECT_ID, require confirmation when the project override disagrees with the SA JSON, and drop unused _is_valid_subscription_id.
…2276)

* import support for nsf records and folders

* added permissions for nsf import

* added sample data

* updated tests

Co-authored-by: pvagare-ks <pvagare@keepersecurity.com>
* Add --format json support to whoami and PAM list/info commands

* reverted changes for whoami and pam action job-list
…evoked. (#2271) (#2279)

* Block PAM workflow admin commands after workflow-settings policy is revoked.

* Fix review comments.
Every authentication branch in the entrypoint ended with an unconditional
`sleep infinity`, so a container invoked with a one-shot command printed its
result and then never exited. That made the image unusable from scripts, since
the command's exit status never propagated, and there was no way to opt out.
Only failing commands exited, and then only incidentally, via `set -e`.

Replace the eight unconditional `sleep infinity` calls with a single lifecycle
decision: stay resident only for service mode, for an invocation with no
command, or when KEEPER_KEEP_ALIVE is set; otherwise exit with the wrapped
command's status.

Three further problems in the same lifecycle code:

- The container ignored SIGTERM. Bash defers trap handlers while a foreground
  child runs, so `sleep infinity` kept the existing EXIT/INT/TERM trap from
  ever firing: `docker stop` waited out the full grace period and then
  SIGKILLed, leaving the KSM config monitor uncleaned. Idle via a backgrounded
  sleep and `wait`, with explicit TERM/INT handlers.

- Command arguments were flattened into a single string and re-split by word
  splitting, so any argument containing spaces (record titles, notes, search
  queries) reached Commander as several arguments. Carry them in an array.

- The KSM one-shot path no longer starts the perpetual config monitor. It
  uploads config.json once so refreshed device state persists, then exits.

Fixes #2264
…ps (#2281) (#2285)

* Require ownership before adopting service Docker setup folders and apps

* Use folder_add instead of mkdir to create duplicate name share folder
The view action previously emitted flattened Name/Value rows even in
JSON format. It now emits the complete record contents in the same
structure as the get command: raw decrypted record data for typed (v3+)
records, and get-style field names (login, password, login_url,
custom_fields, totp, attachments) for legacy records, plus revision
metadata (version, revision, modified_by, client_modified_time). This
allows raw history diffs between revisions. Table output is unchanged.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…record (#2283) (#2287)

* Fix pam rbi edit so NSF RBI settings persist via update_pam_record

* Add NSF rbi edit regression test and debug log for is_pam_nsf_record fallback
@mfordkeeper

Copy link
Copy Markdown
Contributor Author

Replaced with #2295

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

10 participants