fix: prevent stale group credential acceptance blocking issuance

[lenkan]

Problem

A stale multisig credential-issuance acceptance could poison a member while another issuance was in flight:

1. In a 2-of-2 group, M1 issues credential 1 and M2 accepts it successfully.
2. M1 starts credential 2, leaving M2's /multisig/iss notification unread.
3. M2 accepts credential 1's old exchange again. The wallet creates a fresh KERIA issue request from credential 1's completed ACDC; it does not replay the original HTTP request, KEL, TEL, or signatures.
4. M2 then accepts credential 2.

Before this change, step 3 could return HTTP 500 and leave M2 unable to finish or read credential 2. M1 could also remain blocked waiting for M2's contribution.

Related to #213.

Fix

Treat an already-completed group credential acceptance as idempotent by reusing its persisted operation state, without submitting a conflicting KEL or TEL event.

[codecov]

Codecov Report

❌ Patch coverage is 78.21782% with 44 lines in your changes missing coverage. Please review.
✅ Project coverage is 87.41%. Comparing base (f621ca4) to head (ccd3197).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
src/keria/app/credentialing.py	74.69%	42 Missing ⚠️
src/keria/core/longrunning.py	94.44%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #449      +/-   ##
==========================================
- Coverage   87.76%   87.41%   -0.35%     
==========================================
  Files          26       26              
  Lines        5826     5992     +166     
==========================================
+ Hits         5113     5238     +125     
- Misses        713      754      +41     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[lenkan]

@iFergal @kentbull is this something you have experienced?

[iFergal]

@lenkan No, but I am in favour of making the APIs idempotent