fix(acp): forward resolved credential pool into ACP-created agents

[kshitijk4poor]

Summary

ACP-created AIAgent instances were missing the resolved credential_pool, preventing long-lived OpenAI Codex ACP sessions from using Hermes' in-process OAuth 401 refresh/rotation path.

Changes

• acp_adapter/session.py: Forward runtime["credential_pool"] into ACP-created AIAgent instances (parity with gateway's _resolve_runtime_agent_kwargs)
• tests/acp/test_session.py: Extend construction regression to assert the pool is retained

Validation

• 15 ACP session tests pass
• E2E: verified AIAgent.init accepts credential_pool, runtime_provider returns it, gateway already forwards it (this brings ACP to parity)

Credits @stefanpieter for the original fix.

[GottZ]

This was generated by AI during triage.

Summary

Two PRs address #70292 with the same focused change: both forward the resolved credential pool into ACP-created agents and add an identity-based regression assertion, directly fixing the missing pool handoff that disables OAuth recovery in long-lived ACP sessions.

Related pull requests

• fix(acp): forward resolved credential pool #70293 [closed] best fix — (+6/-3) — superseded by fix(acp): forward resolved credential pool into ACP-created agents #77029: The diff forwards runtime.get("credential_pool") in acp_adapter/session.py and verifies sentinel-pool identity in tests/acp/test_session.py. Despite the keep_open review on fix(acp): forward resolved credential pool #70293, contributor kshitijk4poor states that this fix was cherry-picked into fix(acp): forward resolved credential pool into ACP-created agents #77029 with authorship preserved, and the two diffs are substantively identical, so the closed PR remains relevant as the original implementation.
• fix(acp): forward resolved credential pool into ACP-created agents #77029 best fix — (+6/-3) — keep open with a salvage path: Preserve the focused credential_pool forwarding and sentinel identity regression test; together they directly restore the existing credential-pool recovery path for ACP-created agents without broadening authentication behavior.

Duplicates

#70293 and #77029 implement substantially the same one-line ACP credential-pool handoff and the same regression assertion; #70293 is the original, now-closed implementation superseded by #77029.

Suggested consolidation

Keep #77029 open with the concrete salvage path of retaining the focused credential_pool handoff and its identity-based ACP construction regression test. Treat the already-closed #70293 as superseded by and duplicate of #77029; no additional implementation should be carried forward from it because its substantive diff is already present in #77029.

Complex graph

flowchart LR
    classDef open fill:#dbeafe,stroke:#1d4ed8,color:#1e3a8a
    classDef merged fill:#dcfce7,stroke:#15803d,color:#14532d
    classDef closed fill:#e5e7eb,stroke:#6b7280,color:#1f2937
    classDef unverified fill:#f3f4f6,stroke:#9ca3af,color:#374151
    classDef best stroke-width:3px,stroke:#b45309
    classDef target stroke-width:3px,stroke:#4338ca
    I70292(["issue #70292 (open)"])
    subgraph Dup70293 ["PRs duplicating each other"]
        P70293["PR #70293 (closed)"]
        P77029["PR #77029 (open)"]
    end
    P77029 -->|best fix| I70292
    class I70292 open
    class P70293 closed
    class P77029 open
    class P70293 best
    class P77029 best
    class P77029 target
    click I70292 "https://github.com/NousResearch/hermes-agent/issues/70292"
    click P70293 "https://github.com/NousResearch/hermes-agent/pull/70293"
    click P77029 "https://github.com/NousResearch/hermes-agent/pull/77029"

Loading

Graph: solid arrow = fixes / best fix, dashed arrow = partial or unverified (see edge label); boxed group = PRs duplicating each other; amber border = best fix; indigo border = target; gray node = closed (state tag in the node label).

Cross-PR triage: Reviewed 2 pull requests and 1 issue in this complex. Each diff was read against this issue; Assessment working set: 4 kB of PR diffs, 8 kB of issue/PR text, <1 kB of discussion (2 comments), 4 verify verdicts. verdicts reflect diff content, not PR titles. Part of an automated triage batch.