Skip to content
Merged
Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
20 commits
Select commit Hold shift + click to select a range
402c2f2
Foundation catalog drift DD-1/DD-2: one Agents SDK disposition and th…
Oct 3, 2026
226f079
Register the catalog drift files (hot-file protocol, last commit)
Oct 3, 2026
2457bf2
Keep the frozen native-clients verdict fields; reconcile the agent-sd…
Oct 3, 2026
75e661f
Re-register the catalog drift files (hot-file protocol, last commit)
Oct 3, 2026
5ba34a6
Catalog drift review round: authentication claims, carried #15167 hea…
Oct 3, 2026
38e8840
Re-register the catalog drift review-round files (hot-file protocol)
Oct 3, 2026
e5b044a
Merge main 9b0b8d6d2 into the catalog drift branch; the harness-defau…
Oct 3, 2026
eebc701
Merge main 4ced2923 into the catalog drift branch (hot-file protocol:…
Oct 3, 2026
ca9e6d3
Regenerate the new-WSL handbook for the research-state.json change an…
Oct 3, 2026
db4efad
Re-register the regenerated handbook and its receipt (hot-file protoc…
Oct 3, 2026
4916f8b
Merge main d2fc3803e into the catalog drift branch; the new-WSL handb…
Oct 3, 2026
dbcd782
Merge main 1f5a791b0 into the catalog drift branch; harness-defaults …
Oct 3, 2026
d4b9bba
Set manifests/evidence.json to main 1f5a791b0's copy (hot-file protoc…
Oct 3, 2026
89ae84d
Re-register the catalog drift files on main 1f5a791b0 (hot-file proto…
Oct 3, 2026
32ee928
Merge main 1a64e8f3e into the catalog drift branch; the new-WSL handb…
Oct 4, 2026
41d7b24
Set manifests/evidence.json to main 1a64e8f3e's copy (hot-file protoc…
Oct 4, 2026
e0b2815
Re-register the catalog drift files on main 1a64e8f3e (hot-file proto…
Oct 4, 2026
ea503af
Merge main e7c297e25 for the custody merge (hot-file protocol: main's…
Oct 4, 2026
aba524d
Hot-file protocol: reset manifests/evidence.json to the merge base be…
Oct 4, 2026
69686a7
Register this branch's files on the merge-base registry (hot-file pro…
Oct 4, 2026
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions catalogs/foundation/manifest.json
Original file line number Diff line number Diff line change
Expand Up @@ -20,7 +20,7 @@
"id": "native-clients",
"title": "Native clients",
"purpose": "Run ordinary coding, research and review through locally signed-in native clients.",
"selection": "Codex and Claude keep native accounts, model selection, caching and compaction. Model routing and providers are not a separate layer here, and the catalogs/us-equities/models.json ledger remains a note on this native-clients selection. The user decided on 2026-09-27 to adopt a loopback OmniRoute gateway (component omniroute) for pooled GPT-6 access by gateway-routed Codex lanes such as the landscape sweep. It is a source build of release/v3.8.51 with an upstream fix (#14904) and an upstream feature (#13788); see docs/decisions/2026-09-27-omniroute-account-pool.md. Native Codex stays the max-quality default. The agent-sdks layer default stays pending the preregistered three-arm workers comparison (docs/grand-catalog-handbook.md, agent-sdks row), which remains its only acceptance gate. The terminal surface of these clients (tab titles, bell, colour depth, needed-only alerts) is settled by docs/decisions/2026-09-28-terminal-experience.md: AI-client (Claude, Codex) Windows Terminal profiles must not suppress application titles (static shell profiles may keep fixed titles), BEL is the only bell or notification signal Windows Terminal 1.24 and 1.25 act on (OSC 9;4 only sets progress state), and alerts are limited to pending decisions.",
"selection": "Codex and Claude keep native accounts, model selection, caching and compaction. Model routing and providers are not a separate layer here, and the catalogs/us-equities/models.json ledger remains a note on this native-clients selection. The user decided on 2026-09-27 to adopt a loopback OmniRoute gateway (component omniroute) for pooled GPT-6 access by gateway-routed Codex lanes such as the landscape sweep. Dated workstation observations reconciled on 2026-10-03 identify 20128 as omniroute-3.8.51-2f42a9ac-pr13788-affinity2-pr15167 (BUILD_SHA cf6748d04) and 20129 as omniroute-3.8.51-2f42a9ac-pr13788 (BUILD_SHA 87c4c488d). Both use base 2f42a9ac19d1a247ec9ce5473b790843724b3061, whose tree matches v3.8.51. Both carry upstream #13788 (head 6c7990058c4ce9677de79452c8cefb10b4bf1b3d); 20128 also carries the local affinity patch and upstream #15167 (head 0585aba5589d5a1f49243a13a8db249558e7c9e3). The clean omniroute@3.8.51 package caps gpt-6.1-sol at xhigh; upstream PR #15167 keeps requested max (and maps ultra to wire max). Sources: docs/decisions/2026-09-30-omniroute-rebuild.md, PR #637's decision at https://github.com/seathatflowsinourveins/native-agent-stack/blob/269f9516883b/docs/decisions/2026-10-03-omniroute-3851-pin.md, and pinned upstream effort sources linked in docs/decisions/2026-10-03-foundation-catalog-drift.md. These descriptions are dated observations, not new host acceptance. Native Codex stays the max-quality default. The agent-sdks layer default stays pending the preregistered three-arm workers comparison (docs/grand-catalog-handbook.md, agent-sdks row), which remains its only acceptance gate. The terminal surface of these clients (tab titles, bell, colour depth, needed-only alerts) is settled by docs/decisions/2026-09-28-terminal-experience.md: AI-client (Claude, Codex) Windows Terminal profiles must not suppress application titles (static shell profiles may keep fixed titles), BEL is the only bell or notification signal Windows Terminal 1.24 and 1.25 act on (OSC 9;4 only sets progress state), and alerts are limited to pending decisions.",
"activation": "Use the intended native client and inspect its loaded project tools; new process defaults require a fresh session.",
"lifecycle_scope": "Dated native source retrieval, process environment and scoped session continuation passed; host-wide restart recovery is separate.",
"next_gap": "Bind fresh-session activation and resumed effects to the next target host without repeating accepted model trials."
Expand Down Expand Up @@ -164,10 +164,10 @@
"id": "agent-sdks",
"title": "Agent SDKs and runtime workers",
"purpose": "Select the programmatic SDK or runtime that gives a worker structured events, custom tools or session control beyond the native CLI.",
"selection": "Retain native Codex CLI/SDK for session work. Use an explicitly configured OmniRoute SDK lane for application tools/events/handoffs; Pi remains an optional trial. The source-only 2026-10-02 two-host review is in docs/decisions/2026-10-02-two-host-north-star-architecture.md; it does not promote a new runtime.",
"selection": "Retain native Codex CLI/SDK for session work. Use an explicitly configured OmniRoute SDK lane for application tools/events/handoffs; Pi remains an optional trial. OpenAI Agents SDK v0.23.1 is an unqualified alternative: OPENAI_BASE_URL supports an OpenAI-compatible endpoint and its experimental codex_tool can reuse the local Codex login, while the managed Agents API separately requires an API key and bills at API rates. SDK requests through OmniRoute have not been measured; AD-5's codex_tool adoption stays on hold. The source-only 2026-10-02 two-host review is in docs/decisions/2026-10-02-two-host-north-star-architecture.md; it does not promote a new runtime. The pinned SDK sources and 2026-10-03 catalog correction are in docs/decisions/2026-10-03-foundation-catalog-drift.md.",
"activation": "Use the native CLI for ordinary work; reach for an official SDK only when an application needs structured events, custom tools or programmatic session control.",
"lifecycle_scope": "Retain each runtime's recorded host, version and task scope. Native Codex receipts, separate SDK protocol checks and the unmerged Pi trial are distinct; the 2026-10-02 source review supplies no new model, runtime or workstation lifecycle acceptance.",
"next_gap": "The 2026-10-02 source review distinguishes native Ultra from API reasoning enums and Pi 0.99.1 trial evidence from upstream 1.0.0. Qualify the actual application's endpoint, locked SDK, extensions, cancellation and resume before broader promotion; workstation activation remains host-specific."
"next_gap": "The 2026-10-02 source review distinguishes native Ultra from API reasoning enums and Pi 0.99.1 trial evidence from upstream 1.0.0. Qualify the actual application's endpoint, locked SDK, extensions, cancellation and resume before broader promotion; workstation activation remains host-specific. AD-5 remains a hold: OpenAI Agents SDK codex_tool adoption needs a demonstrated orchestration requirement, a frozen comparison against the Codex thread API and codex exec --json plus resume with complete usage and effect/recovery checks, and graduation from experimental."
Comment thread
seathatflowsinourveins marked this conversation as resolved.
Resolved
},
{
"id": "mcp-surfaces",
Expand Down
39 changes: 30 additions & 9 deletions catalogs/landscape/foundation.json
Original file line number Diff line number Diff line change
Expand Up @@ -46,12 +46,15 @@
{
"name": "OpenAI Agents SDK",
"repository": "https://github.com/openai/openai-agents-python",
"disposition": "conditional",
"rationale": "Fits an application-owned API agent; it is not a substitute for the established subscription-backed native Codex account and behavior.",
"disposition": "unqualified",
"rationale": "Source review corrected on 2026-10-03. The managed Agents API requires OPENAI_API_KEY and bills at API rates. The open-source SDK supports OpenAI-compatible endpoints through OPENAI_BASE_URL; its experimental codex_tool launches the local Codex CLI and can use the existing Codex login with the inherited environment when no API key override is passed. SDK requests through OmniRoute have not been measured. AD-5 remains a hold: no demonstrated orchestration requirement or accepted three-arm comparison, and codex_tool is still experimental.",
Comment thread
seathatflowsinourveins marked this conversation as resolved.
Resolved
"evidence_kind": "source_review",
"evidence_refs": [
"docs/foundation-closure-20260921.md",
"catalogs/us-equities/agents-operations.json"
"docs/decisions/2026-10-03-foundation-catalog-drift.md",
"https://developers.openai.com/api/docs/guides/agents-api/overview",
"https://github.com/openai/openai-agents-python/blob/v0.23.1/src/agents/extensions/experimental/codex/exec.py#L226-L240",
"https://github.com/openai/openai-agents-python/blob/v0.23.1/docs/config.md#L84-L88",
"https://github.com/openai/openai-agents-python/blob/v0.23.1/docs/tools.md#L890-L894"
]
},
{
Expand Down Expand Up @@ -178,10 +181,14 @@
"name": "OpenAI Agents SDK",
"repository": "https://github.com/openai/openai-agents-python",
"disposition": "unqualified",
"why_not_default": "The only evidence is source review. catalogs/us-equities/agents-operations.json line 1611 describes it as API-backed. Line 1624 says 'Prospective commands; not executed'. Lines 1634-1635 say 'No provider API call or model comparison performed' and that it 'Does not inherit Codex account entitlement'. Because it is API-backed, it does not meet the requirement to use existing native accounts. Correction from the earlier round: docs/foundation-closure-20260921.md is dropped from the refs because it does not mention this SDK.",
"why_not_default": "Source review corrected on 2026-10-03. The managed Agents API requires OPENAI_API_KEY and bills at API rates. The open-source SDK supports OpenAI-compatible endpoints through OPENAI_BASE_URL; its experimental codex_tool launches the local Codex CLI and can use the existing Codex login with the inherited environment when no API key override is passed. SDK requests through OmniRoute have not been measured. AD-5 remains a hold: no demonstrated orchestration requirement or accepted three-arm comparison, and codex_tool is still experimental. The 2026-09-21 SDK table has no OpenAI Agents SDK row; the older API-backed premise does not establish blanket native-account incompatibility.",
"evidence_class": "source_review",
"evidence_refs": [
"catalogs/us-equities/agents-operations.json"
"docs/decisions/2026-10-03-foundation-catalog-drift.md",
"https://developers.openai.com/api/docs/guides/agents-api/overview",
"https://github.com/openai/openai-agents-python/blob/v0.23.1/src/agents/extensions/experimental/codex/exec.py#L226-L240",
"https://github.com/openai/openai-agents-python/blob/v0.23.1/docs/config.md#L84-L88",
"https://github.com/openai/openai-agents-python/blob/v0.23.1/docs/tools.md#L890-L894"
],
"source": "lane:claude"
},
Expand Down Expand Up @@ -213,7 +220,8 @@
"native Claude Code CLI (2.1.278 or current pin)",
"native Codex CLI/app-server (0.155.1 or current pin)",
"Claude Agent SDK on the same account and fixture",
"Codex SDK on the same account and fixture"
"Codex SDK on the same account and fixture",
"OpenAI Agents SDK codex_tool on the same Codex login and fixture, only after the AD-5 orchestration requirement and experimental-status hold can be revisited"
],
"fixture_paths": [
"blueprints/convergence-practice/native-recovery/claude/README.md",
Expand Down Expand Up @@ -252,7 +260,7 @@
"MCPorter's referenced readiness output includes a Context Mode daemon error; version and listing success do not establish functional bridge recovery."
],
"checked_at": "2026-09-22",
"verdict_overturn_when": "Change the verdict only on an executed comparison, not on source review. The condition has two parts. First, the native path cannot deliver a structured-event, custom-tool or programmatic-session capability that an application needs. Second, an SDK arm (Claude Agent SDK or Codex SDK) passes the same frozen recovery fixture on the same account. The fixtures are blueprints/convergence-practice/native-recovery/claude/ (checkpoint, SIGINT, same-session finalize) and blueprints/convergence-practice/native-recovery/ (Codex interrupt and same-thread resume). The SDK arm must match the receipt checks: same session/thread ID, checkpoint unchanged and checkpoint executed once. The offline oracles must still pass: 'python3 -m unittest tests.test_native_recovery' and 'python3 -m unittest discover -s blueprints/convergence-practice/native-recovery/claude -p test_*.py'. A separate trigger: either native client fails a rerun of the same-prompt read-only research task recorded in evidence/artifacts/full-stack-convergence-20260921/native-client-results.json on the intended host. That task has 11 tool results and no tool errors."
"verdict_overturn_when": "Change the verdict only on an executed comparison, not on source review. The condition has two parts. First, the native path cannot deliver a structured-event, custom-tool or programmatic-session capability that an application needs. Second, an SDK arm (Claude Agent SDK, Codex SDK or OpenAI Agents SDK codex_tool) passes the same frozen recovery fixture on the same account. OpenAI Agents SDK codex_tool remains held under AD-5 until a demonstrated orchestration requirement, graduation from experimental and an accepted frozen three-arm comparison with complete usage and effect/recovery checks justify adoption; see docs/decisions/2026-10-03-foundation-catalog-drift.md. The fixtures are blueprints/convergence-practice/native-recovery/claude/ (checkpoint, SIGINT, same-session finalize) and blueprints/convergence-practice/native-recovery/ (Codex interrupt and same-thread resume). The SDK arm must match the receipt checks: same session/thread ID, checkpoint unchanged and checkpoint executed once. The offline oracles must still pass: 'python3 -m unittest tests.test_native_recovery' and 'python3 -m unittest discover -s blueprints/convergence-practice/native-recovery/claude -p test_*.py'. A separate trigger: either native client fails a rerun of the same-prompt read-only research task recorded in evidence/artifacts/full-stack-convergence-20260921/native-client-results.json on the intended host. That task has 11 tool results and no tool errors."
},
{
"catalog": "foundation",
Expand Down Expand Up @@ -4558,7 +4566,7 @@
"requirement": "Give a worker structured events, custom tools or session control through a supported SDK when the native CLI's bounded review is not enough.",
"current_choice": "Codex SDK/CLI (component codex) for supported start/run/resume; Claude Agent SDK, OpenHands SDK, Temporal and LangGraph remain source-reviewed, unselected alternatives.",
"decision": "retain",
"rationale": "Codex SDK integration is accepted only within its recorded tests and needs no new runtime; the SDK table in docs/foundation-closure-20260921.md (lines 78-84) reviewed the alternatives on 2026-09-21 and none has since been revisited by any lane.",
"rationale": "Codex SDK integration is accepted only within its recorded tests and needs no new runtime. The five rows of the 2026-09-21 SDK table in docs/foundation-closure-20260921.md are at lines 82-86 and do not include OpenAI Agents SDK. The 2026-10-03 pinned-source correction adds that SDK as unqualified without qualifying its experimental codex_tool or its requests through OmniRoute; see docs/decisions/2026-10-03-foundation-catalog-drift.md.",
"evidence_refs": [
"docs/foundation-closure-20260921.md",
"catalogs/foundation/decisions.json"
Expand All @@ -4580,6 +4588,19 @@
"docs/foundation-closure-20260921.md"
]
},
{
"name": "OpenAI Agents SDK",
"repository": "https://github.com/openai/openai-agents-python",
"disposition": "unqualified",
"rationale": "Source review at v0.23.1 supports an OpenAI-compatible SDK endpoint and an experimental codex_tool that can inherit the existing local Codex login; the managed Agents API separately requires an API key and bills at API rates. SDK requests through OmniRoute have not been measured. AD-5 stays on hold until a needed orchestration capability, an accepted frozen comparison against the Codex thread API and codex exec --json plus resume, complete usage and effect/recovery checks, and graduation from experimental justify adoption.",
Comment thread
seathatflowsinourveins marked this conversation as resolved.
Resolved
"evidence_kind": "source_review",
"evidence_refs": [
"docs/decisions/2026-10-03-foundation-catalog-drift.md",
"https://github.com/openai/openai-agents-python/blob/v0.23.1/src/agents/extensions/experimental/codex/exec.py#L226-L240",
"https://github.com/openai/openai-agents-python/blob/v0.23.1/docs/config.md#L84-L88",
"https://github.com/openai/openai-agents-python/blob/v0.23.1/docs/tools.md#L890-L894"
]
},
{
"name": "Claude Agent SDK",
"repository": "https://github.com/anthropics/claude-agent-sdk-python",
Expand Down
2 changes: 1 addition & 1 deletion catalogs/us-equities/decision-index.json
Original file line number Diff line number Diff line change
Expand Up @@ -15024,7 +15024,7 @@
"kind": "research_supplement",
"path": "catalogs/landscape/foundation.json",
"pointer": "/layers/0/candidates/2",
"disposition": "conditional"
"disposition": "unqualified"
},
{
"kind": "research_supplement",
Expand Down
Loading
Loading