mirror of
https://github.com/dtzp555-max/olp.git
synced 2026-07-19 09:45:07 +00:00
fix+release(v0.4.2): D75 — codex CLI v0.133.0 schema + per-hop model override (#47)
Patch release fixing 5 bugs caught by real-machine E2E testing on PI231 +
Mac mini (2026-05-26 session). Prior D-day reviewers + the post-v0.4.0
maintainer review all missed these because they reviewed against spec text
and against the local OLP install's cached codex CLI shape, not against a
fresh `npm install -g @openai/codex` on a remote operator host getting
v0.133.0 for the first time.
F1 — codex auth.json schema pin (lib/providers/codex.mjs readAuthArtifact)
Real codex CLI v0.133.0 nests the access token under `tokens.access_token`,
not at top-level access_token / token / accessToken. Pre-D75 readAuthArtifact
returned null → OLP reported "auth artifact missing" even for fully
logged-in users. Fix: prepend creds?.tokens?.access_token to the precedence
chain at both override + default branches. Legacy fields preserved as
fallback. Authority: codex CLI v0.133.0 on-disk auth.json shape verified
empirically on PI231 2026-05-26 E2E session.
F2 — codex spawn args + --skip-git-repo-check (lib/providers/codex.mjs irToCodex)
codex CLI v0.133.0 trusted-directory sandbox refuses with "Not inside a
trusted directory" outside git repos. OLP deploys typically outside a git
repo. Fix: add '--skip-git-repo-check' to args before '--model'. Authority:
codex CLI v0.133.0 reference (`codex exec --help` documents the flag).
F3 — codex NDJSON event shape pin (lib/providers/codex.mjs codexChunkToIR)
Real v0.133.0 stream: thread.started → turn.started → item.completed
(item.type='agent_message', item.text=<response>) → turn.completed.
D6 defensive parser only recognised top-level content/delta/text +
type:'stop'/done:true → every chunk silently dropped → response body had
content: null. Fix: add three new recognisers (item.completed → delta;
turn.completed → stop; turn.failed → error) before the legacy fallback
chain. Legacy recognisers preserved for backward/forward compat.
F4 — `olp status` reads body.stats.cache.size, not body.cache.entries
(bin/olp.mjs cmdStatus). Server payload nests stats under stats.cache;
CacheStore.stats() exposes {hits, misses, size, inflightCount} — there is
no `entries` field. D74 P2-3 fixed cmdUsage + cmdCache for the same bug
class but missed cmdStatus.
F7 — per-hop chain `model` overrides IR model in provider.spawn()
(server.mjs executeHopFn + streaming sourceFactory). Pre-D75 executeHopFn
used hopModel for cache key + audit ctx but passed the original irReq
(with irReq.model = user's request) to provider.spawn(). Chain config
[{anthropic, claude-X}, {openai, gpt-5.5}] would always spawn BOTH plugins
with --model claude-X — openai rejected the unknown model and the chain
died. This broke the core OLP value prop (cross-provider fallback with
provider-appropriate model substitution). Fix: build per-hop IR variant
with { ...irReq, model: hopModel } and pass to spawn. Conditional skips
clone when hopModel === irReq.model. Applied to BOTH buffered path AND
streaming path. Authority: ADR 0004 § Chain advancement step 1 (per-hop
config supplies provider AND model — contract always specified, code
didn't complete it).
Out of scope (deferred to Phase 5):
- F5 (server bind / OLP_BIND env) — needs anonymous-key trust review
- F6 (doctor client-vs-server-side limit) — needs trigger-taxonomy ADR
Test count: 704 → 714 (+10 Suite 36 D75 regression tests: 36i–36r).
Files touched: lib/providers/codex.mjs, bin/olp.mjs, server.mjs,
test-features.mjs, package.json, CHANGELOG.md.
Phase 5 process learning: every provider plugin D-day must include a
real-CLI E2E on a remote operator host before merging — not on the
maintainer workstation (which may have an older CLI cached from a prior
install). D6/D7 codex E2E was deferred and that deferral compounded across
3 layers. F7 reinforces a separate lesson: when a function signature takes
(provider, model, ir), reviewers must check that `model` is consumed
everywhere downstream — not just at the call site they happened to look at.
Authority: ADR 0002 (provider contract — codex plugin), ADR 0004 (fallback
engine — per-hop model contract), lib/providers/codex.mjs D6 assumption
A2/A3/A4 docstrings (which all said "D7 will pin" and D7 never did); codex
CLI v0.133.0 on-disk schema + `codex exec --help` output verified
empirically on PI231 (2026-05-26 E2E session); Iron Rule 第二律
evidence-over-should-work; CLAUDE.md release_kit.phase_rolling_mode
cross-Phase discipline ("hotfix to a shipped Phase N deliverable → bump
patch, tag, release before next push").
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
This commit is contained in:
dtzp555-max
@@ -6,6 +6,30 @@ All notable changes to OLP land here. Per `CLAUDE.md` release_kit overlay, this
|
||||
|
||||
(empty — Phase 5 entries land here once Phase 5 opens)
|
||||
|
||||
## v0.4.2 — 2026-05-26
|
||||
|
||||
### Post-v0.4.1 hotfix batch (D75) — real-machine E2E findings
|
||||
|
||||
Patch release fixing 5 bugs caught by **real-machine E2E testing on PI231 + Mac mini (2026-05-26 session)** — bugs that prior D-day reviewers AND the post-v0.4.0 maintainer review both missed because they reviewed against spec text and against the local OLP install's `~/.codex/auth.json` shape (cached from an older codex CLI version), not against real provider CLIs running on a remote operator host that did `npm install -g @openai/codex` for the first time on 2026-05-26 and got codex CLI v0.133.0.
|
||||
|
||||
**Root cause of the missed-bug class.** D6 (codex plugin authoring) explicitly documented three unpinned assumptions (A3 = access-token field name, A4 = NDJSON event schema, A2-adjacent = trusted-directory sandbox). D6 noted "D7 E2E will pin." D7 then shipped without performing real-codex-CLI E2E (the E2E gating mark was carried but the actual run was deferred). Every subsequent D-day reviewer trusted the D6/D7 codex plugin code unchanged because the static review couldn't see that the v0.133.0 CLI had moved the auth-token field, the event schema, AND added a new trusted-directory sandbox flag. The D74 maintainer review focused on `/health` / `/cache/stats` / `/v0/management/dashboard-data` payload shapes — none of which exercise the codex plugin's spawn path. F7 (per-hop model override) is a different class of miss — every reviewer read `executeHopFn(provider, model, ir)` and saw `model` consumed for cache key + audit ctx, but none traced through to confirm `model` is ALSO substituted into the IR passed to `provider.spawn()`. The function signature implied per-hop semantics that the body never fully delivered.
|
||||
|
||||
- **[F1] codex auth.json schema pin — codex CLI v0.133.0 nests the access token under `tokens.access_token`** (verified empirically on PI231 / Mac mini, 2026-05-26). Pre-D75 `readAuthArtifact()` read only top-level `creds.access_token` / `creds.token` / `creds.accessToken` — all undefined under v0.133.0 → returned `null` → OLP reported "auth artifact missing" via `/health` and `olp doctor` AND refused to spawn codex even when the user had fully completed `codex login`. Fix: prepend `creds?.tokens?.access_token` to the precedence chain at BOTH call sites (`OPENAI_CODEX_AUTH_PATH` override branch + default `$CODEX_HOME/auth.json` branch). Legacy top-level fields preserved as fallback for backward compat with older codex CLI versions.
|
||||
- **[F2] codex spawn args — codex CLI v0.133.0 trusted-directory sandbox requires `--skip-git-repo-check`.** v0.133.0 refuses with `"Not inside a trusted directory and --skip-git-repo-check was not specified."` when spawned outside a git repo, exits non-zero with zero NDJSON output → OLP surfaces `SPAWN_FAILED` with no usable chunks → the fallback engine advances to next hop unnecessarily even when codex is configured and authenticated. OLP's typical deploy CWD (`~/olp/`) is NOT a git repo on operator hosts. Fix: add `'--skip-git-repo-check'` to the args array before `--model`. OLP is the trusted caller (operator's own server invoking the operator's own subscription via documented `codex exec` automation); the sandbox safeguards interactive shells, not pre-authorized automation.
|
||||
- **[F3] codex NDJSON event shape pin — codex CLI v0.133.0 emits `item.completed` + `turn.completed` + `turn.failed`**, not the D6-assumed `content`/`delta`/`text` + `type:'stop'`/`done:true` shapes. Real v0.133.0 stream (verified empirically): `{"type":"thread.started",...}` → `{"type":"turn.started"}` → `{"type":"item.completed","item":{"id":"item_0","type":"agent_message","text":"<response>"}}` → `{"type":"turn.completed","usage":{...}}`. Pre-D75, every chunk was silently dropped by `codexChunkToIR()` → response body had `content: null`. Fix: prepend three new recognizers (`item.completed` with `item.type === 'agent_message'` → IR delta; `turn.completed` → IR stop; `turn.failed` → IR error). Legacy D6 defensive recognizers preserved below as forward/backward compat fallbacks.
|
||||
- **[F4] `olp status` reads `body.stats.cache.size` (not OCP-era `body.cache.entries`).** Same class as D74 P2-3 (which fixed `cmdUsage` + `cmdCache`); D74 missed the parallel bug in `cmdStatus`. Server payload nests cache stats as `body.stats.cache.{hits, misses, size, inflightCount}` per `server.mjs handleManagementStatus`, and `CacheStore.stats()` has no `entries` field per `lib/cache/store.mjs`. Pre-D75 output showed `entries=?`. Fix: read `c.size` for entries display; also surface `inflightCount` when present.
|
||||
- **[F7] per-hop chain `model` field now overrides IR model in `provider.spawn()`.** Pre-D75 `executeHopFn(hopProvider, hopModel, irReq)` used `hopModel` for cache key + audit ctx but passed the ORIGINAL `irReq` (with `irReq.model` = user's original request) to `hopProviderPlugin.spawn(irReq, authContext)`. A chain config `[{provider:anthropic, model:claude-X}, {provider:openai, model:gpt-5.5}]` would always spawn BOTH plugins with `--model claude-X` — openai rejected the unknown model and the chain died. This broke the core OLP value prop (cross-provider fallback with provider-appropriate model substitution per hop). Fix: build a per-hop IR variant with `{...irReq, model: hopModel}` and pass that to spawn. Conditional skips clone when `hopModel === irReq.model` (common case: single-provider chains, or single-hop chains where the chain config repeats the request model). Applied to BOTH the buffered path (`executeHopFn`) AND the streaming path (`sourceFactory` for `getOrComputeStreaming`). **Authority:** ADR 0004 § Chain advancement step 1 (per-hop config supplies provider AND model — the contract was always specified, but the code didn't complete it).
|
||||
|
||||
**Phase 5 process learning recorded.** Every provider plugin's D-day must include a real-CLI E2E ON A REMOTE OPERATOR HOST before merging — not on the maintainer workstation (which may have an older CLI cached from a prior install, hiding new field renames / new sandbox flags / new event shapes). The D6/D7 codex E2E was deferred and that deferral compounded across 3 layers (D6 = unpinned, D7 = pinning deferred, D8+ = trusted D6/D7 unchanged). F7 reinforces a separate lesson: when a function signature takes `(provider, model, ir)`, reviewers must check that `model` is consumed everywhere downstream, not just at the call site they happened to look at.
|
||||
|
||||
**Out of D75 scope (deferred to Phase 5 explicit ADR amendments):**
|
||||
- F5 (server bind 127.0.0.1 / `OLP_BIND` env) — needs `lib/keys.mjs` anonymous-key trust boundary review before binding to non-loopback by default
|
||||
- F6 (`olp doctor` client-vs-server-side limit detection) — needs design ADR amendment for trigger taxonomy
|
||||
|
||||
- **Test count delta:** 704 (v0.4.1) → 714 (v0.4.2). +10 D75 regression tests in Suite 36 (36i through 36r).
|
||||
- **Files touched:** `lib/providers/codex.mjs` (F1+F2+F3), `bin/olp.mjs` (F4 cmdStatus), `server.mjs` (F7 buffered + streaming spawn sites), `test-features.mjs` (Suite 36 extension), `package.json` (version), `CHANGELOG.md` (this entry).
|
||||
- **Authority:** ADR 0002 (provider contract — codex plugin), ADR 0004 (fallback engine — per-hop model contract), `lib/providers/codex.mjs` D6 assumption A2/A3/A4 docstrings (which all said "D7 will pin" and D7 never did); codex CLI v0.133.0 on-disk schema + `codex exec --help` output verified empirically on PI231 (2026-05-26 E2E session); Iron Rule 第二律 evidence-over-should-work; CLAUDE.md `release_kit.phase_rolling_mode` cross-Phase discipline.
|
||||
|
||||
## v0.4.1 — 2026-05-26
|
||||
|
||||
### Post-Phase-4 hotfix batch (D74) — maintainer-review findings
|
||||
|
||||
+7
-1
@@ -251,9 +251,15 @@ async function cmdStatus(flags, io) {
|
||||
}
|
||||
io.log(` total reqs: ${body.stats?.total_requests ?? 0}`);
|
||||
io.log(` active reqs: ${body.stats?.active_requests ?? 0}`);
|
||||
// D75 F4 fix: server payload nests cache stats under stats.cache (per
|
||||
// server.mjs handleManagementStatus, ~line 2092). The CacheStore.stats()
|
||||
// contract returns { hits, misses, size, inflightCount } per
|
||||
// lib/cache/store.mjs — there is no `entries` field. Pre-D75 cmdStatus read
|
||||
// `body.stats.cache.entries` (OCP-era) which was always undefined → output
|
||||
// showed "entries=?". Same pattern as D74 P2-3 applied to cmdCache/cmdUsage.
|
||||
if (body.stats?.cache) {
|
||||
const c = body.stats.cache;
|
||||
io.log(` cache: hits=${c.hits ?? 0} misses=${c.misses ?? 0} entries=${c.entries ?? '?'}`);
|
||||
io.log(` cache: hits=${c.hits ?? 0} misses=${c.misses ?? 0} entries=${c.size ?? 0}${typeof c.inflightCount === 'number' ? ` inflight=${c.inflightCount}` : ''}`);
|
||||
}
|
||||
if (Array.isArray(body.recent_errors) && body.recent_errors.length > 0) {
|
||||
io.log(` recent errors: ${body.recent_errors.length}`);
|
||||
|
||||
+108
-3
@@ -157,6 +157,36 @@ function resolveCodexBin() {
|
||||
// D6 assumption A2: auth file is named auth.json (unconfirmed — D7 will pin).
|
||||
// D6 assumption A3: access token field is `access_token` or `token` (unconfirmed).
|
||||
//
|
||||
// ── D75 (v0.4.2) F1 — codex CLI v0.133.0 schema pin ─────────────────────────
|
||||
//
|
||||
// Real codex CLI v0.133.0 auth.json (verified empirically on PI231 / Mac mini,
|
||||
// 2026-05-26 E2E session):
|
||||
//
|
||||
// {
|
||||
// "auth_mode": "chatgpt",
|
||||
// "OPENAI_API_KEY": null | "<key>",
|
||||
// "tokens": {
|
||||
// "id_token": "<JWT>",
|
||||
// "access_token": "<opaque-or-JWT>", <-- THIS is the access token
|
||||
// "refresh_token": "<opaque>",
|
||||
// "account_id": "<uuid>"
|
||||
// },
|
||||
// "last_refresh": "<ISO8601>"
|
||||
// }
|
||||
//
|
||||
// D6 assumption A3 originally tried `creds.access_token` at TOP level. Under
|
||||
// codex v0.133.0 this field does not exist at the top level → readAuthArtifact()
|
||||
// returned null even when the user had fully completed `codex login`. OLP then
|
||||
// reported "auth artifact missing" via /health and `olp doctor`, and refused
|
||||
// to spawn codex — false negative blocking the entire openai provider.
|
||||
//
|
||||
// Fix: prepend `creds?.tokens?.access_token` to the precedence chain. Keep all
|
||||
// existing fallbacks unchanged so older codex CLI versions (pre-v0.133) and any
|
||||
// future shape variants still resolve.
|
||||
//
|
||||
// Authority pin: codex CLI v0.133.0 source + on-disk auth.json captured during
|
||||
// PI231 E2E. See D75 commit body for the verification transcript.
|
||||
//
|
||||
// Returns { accessToken: string } or null (never throws).
|
||||
export function readAuthArtifact() {
|
||||
// 1. Explicit test override — always takes precedence.
|
||||
@@ -165,7 +195,12 @@ export function readAuthArtifact() {
|
||||
try {
|
||||
const raw = readFileSync(authPathOverride, 'utf8');
|
||||
const creds = JSON.parse(raw);
|
||||
const token = creds?.access_token ?? creds?.token ?? creds?.accessToken;
|
||||
// D75 F1: codex CLI v0.133.0 nests the token under `tokens.access_token`.
|
||||
// Preserve top-level fallbacks for backward / forward compat.
|
||||
const token = creds?.tokens?.access_token
|
||||
?? creds?.access_token
|
||||
?? creds?.token
|
||||
?? creds?.accessToken;
|
||||
if (token && typeof token === 'string') return { accessToken: token };
|
||||
} catch { /* fall through */ }
|
||||
return null; // explicit path set but file missing / malformed
|
||||
@@ -178,8 +213,12 @@ export function readAuthArtifact() {
|
||||
try {
|
||||
const raw = readFileSync(authPath, 'utf8');
|
||||
const creds = JSON.parse(raw);
|
||||
// D6 assumption A3: try common OAuth field names in precedence order.
|
||||
const token = creds?.access_token ?? creds?.token ?? creds?.accessToken;
|
||||
// D75 F1: codex CLI v0.133.0 nests the token under `tokens.access_token`.
|
||||
// Try the nested location FIRST, then fall back to legacy top-level fields.
|
||||
const token = creds?.tokens?.access_token
|
||||
?? creds?.access_token
|
||||
?? creds?.token
|
||||
?? creds?.accessToken;
|
||||
if (token && typeof token === 'string') return { accessToken: token };
|
||||
} catch { /* file missing or malformed */ }
|
||||
|
||||
@@ -242,9 +281,30 @@ export function irToCodex(irRequest) {
|
||||
// model string (e.g., gpt-5.5, gpt-5.4, gpt-5.3-codex).
|
||||
// PROMPT: "Initial instruction for the task. Use '-' to pipe the prompt
|
||||
// from stdin."
|
||||
//
|
||||
// ── D75 (v0.4.2) F2 — codex CLI v0.133.0 trusted-directory sandbox ─────────
|
||||
// codex CLI v0.133.0 added a trusted-directory sandbox: invocations outside
|
||||
// a git repo (or outside any directory explicitly trusted via
|
||||
// `codex config trusted-directories`) refuse with:
|
||||
// "Not inside a trusted directory and --skip-git-repo-check was not specified."
|
||||
// and exit non-zero with zero NDJSON output → OLP surfaces SPAWN_FAILED with
|
||||
// no usable chunks → fallback engine advances to next hop unnecessarily.
|
||||
//
|
||||
// The CWD that OLP spawns from is typically the server install dir (`~/olp/`
|
||||
// on Pi231) which is a git repo on maintainer workstations but is NOT a git
|
||||
// repo on most operator hosts. We bypass the sandbox unconditionally because
|
||||
// OLP is the trusted caller (it is the operator's own server invoking its own
|
||||
// configured Codex subscription via the documented `codex exec` automation
|
||||
// entry point). The trusted-directory sandbox is a foot-gun safeguard for
|
||||
// interactive users; OLP's spawn is non-interactive and pre-authorized.
|
||||
//
|
||||
// Authority: codex CLI v0.133.0 release notes / `codex exec --help` output
|
||||
// documenting `--skip-git-repo-check`. Verified empirically on PI231 E2E
|
||||
// 2026-05-26.
|
||||
const args = [
|
||||
'exec',
|
||||
'--json',
|
||||
'--skip-git-repo-check',
|
||||
'--model', irRequest.model,
|
||||
];
|
||||
|
||||
@@ -291,6 +351,51 @@ export function codexChunkToIR(rawNDJSONLine) {
|
||||
|
||||
if (!event || typeof event !== 'object') return null;
|
||||
|
||||
// ── D75 (v0.4.2) F3 — codex CLI v0.133.0 event shape pin ─────────────────
|
||||
// Real codex CLI v0.133.0 NDJSON event stream (verified empirically on PI231
|
||||
// / Mac mini, 2026-05-26 E2E session):
|
||||
// {"type":"thread.started","thread_id":"019e..."}
|
||||
// {"type":"turn.started"}
|
||||
// {"type":"item.started","item":{"id":"item_0","type":"reasoning","text":""}}
|
||||
// {"type":"item.completed","item":{"id":"item_0","type":"agent_message","text":"<response>"}}
|
||||
// {"type":"turn.completed","usage":{"input_tokens":..,"output_tokens":..}}
|
||||
//
|
||||
// The D6 defensive parser recognized `content`/`delta`/`text` fields at the
|
||||
// top level and `type === 'stop'`/`done === true`. None of these match
|
||||
// v0.133.0's actual shape → every chunk was silently dropped → response body
|
||||
// had `content: null`. F3 adds three NEW recognizers (item.completed →
|
||||
// agent_message; turn.completed → stop; turn.failed → error) BEFORE the
|
||||
// legacy fallback chain. Legacy recognizers preserved for forward/backward
|
||||
// compat (older codex versions; future shape variants).
|
||||
|
||||
// F3-a: agent_message item completion.
|
||||
// codex v0.133.0 emits assistant text as a single item.completed event whose
|
||||
// item.type is 'agent_message' and item.text carries the full text. There
|
||||
// are no incremental deltas — the entire response arrives in one chunk.
|
||||
if (event.type === 'item.completed'
|
||||
&& event.item?.type === 'agent_message'
|
||||
&& typeof event.item?.text === 'string') {
|
||||
return { type: 'delta', content: event.item.text };
|
||||
}
|
||||
|
||||
// F3-b: turn completion → stop chunk.
|
||||
// codex v0.133.0 emits turn.completed with a usage block when the model
|
||||
// finishes. We map this to IR stop with finish_reason 'stop'.
|
||||
if (event.type === 'turn.completed') {
|
||||
return { type: 'stop', finish_reason: 'stop' };
|
||||
}
|
||||
|
||||
// F3-c: turn failure → error chunk.
|
||||
// codex v0.133.0 emits turn.failed with an embedded error object when the
|
||||
// turn cannot complete. Extract a human-readable message for the IR error.
|
||||
if (event.type === 'turn.failed') {
|
||||
const errMsg = (typeof event.error === 'string')
|
||||
? event.error
|
||||
: (event.error?.message ?? 'codex turn.failed');
|
||||
return { type: 'error', error: errMsg };
|
||||
}
|
||||
|
||||
// ── Legacy/fallback recognizers (kept for backward + forward compat) ────
|
||||
// Error event: type === 'error' or error field present
|
||||
// A4: defensive — error shape unconfirmed; D7 will pin actual field names
|
||||
if (event.type === 'error' || (event.error && typeof event.error === 'string')) {
|
||||
|
||||
+1
-1
@@ -1,6 +1,6 @@
|
||||
{
|
||||
"name": "olp",
|
||||
"version": "0.4.1",
|
||||
"version": "0.4.2",
|
||||
"description": "Personal multi-provider LLM proxy. Successor to OCP. One HTTP endpoint, multiple subscriptions behind it, automatic routing + fallback + caching.",
|
||||
"type": "module",
|
||||
"main": "server.mjs",
|
||||
|
||||
+25
-6
@@ -1224,13 +1224,25 @@ async function handleChatCompletions(req, res) {
|
||||
}
|
||||
|
||||
const chunks = [];
|
||||
// try/finally: releaseSpawn MUST fire on every exit path — success
|
||||
// (return at end), spawn throw (caught and re-thrown below), or the
|
||||
// D16 truncation-salvage return. The finally is the only mechanism
|
||||
// that guarantees release across all three.
|
||||
// D75 (v0.4.2) F7 — per-hop model override.
|
||||
// The chain config (routing.chains[<requested-model>][<hop>].model) is the
|
||||
// model name to pass to THIS hop's provider plugin — NOT the model the
|
||||
// user originally requested. Pre-D75, executeHopFn used hopModel for the
|
||||
// cache key + audit ctx but passed the ORIGINAL irReq (with irReq.model
|
||||
// still set to the user's request) into hopProviderPlugin.spawn(). Result:
|
||||
// a 2-hop chain like [{anthropic, claude-sonnet-4-6}, {openai, gpt-5.5}]
|
||||
// would spawn codex with --model claude-sonnet-4-6 on hop 1 — openai rejects
|
||||
// the unknown model and the chain dies. This broke the core OLP value prop
|
||||
// (cross-provider fallback with provider-appropriate model substitution).
|
||||
//
|
||||
// Fix: build a per-hop IR variant with the hop's model substituted. Skip
|
||||
// the clone when hopModel === irReq.model (single-provider chains and
|
||||
// chain hops whose model matches the request). Authority: ADR 0004 §
|
||||
// Chain advancement step 1 (per-hop config supplies provider AND model).
|
||||
const hopIrReq = irReq.model === hopModel ? irReq : { ...irReq, model: hopModel };
|
||||
try {
|
||||
try {
|
||||
for await (const irChunk of hopProviderPlugin.spawn(irReq, authContext)) {
|
||||
for await (const irChunk of hopProviderPlugin.spawn(hopIrReq, authContext)) {
|
||||
// D16: check error chunks BEFORE pushing — preserves the invariant that
|
||||
// chunks array contains only delta/stop chunks. Without this, the catch
|
||||
// block's `chunks.length > 0` would mistake a single error chunk for
|
||||
@@ -1422,9 +1434,16 @@ async function handleChatCompletions(req, res) {
|
||||
// releaseSpawn fires exactly once in finally — regardless of normal
|
||||
// exhaustion, mid-stream throw, or iterator.return() from cache-layer
|
||||
// sourceAbortController propagation (§9).
|
||||
//
|
||||
// D75 (v0.4.2) F7 — per-hop model override (streaming path).
|
||||
// Mirror the buffered-path fix: pass the hop's configured model (streamModel)
|
||||
// into streamPlugin.spawn(), not the user's original ir.model. Skip the
|
||||
// clone when ir.model === streamModel. See executeHopFn() above for the
|
||||
// full F7 rationale + authority citation.
|
||||
const streamIr = ir.model === streamModel ? ir : { ...ir, model: streamModel };
|
||||
return (async function* sourceWithRelease() {
|
||||
try {
|
||||
for await (const irChunk of streamPlugin.spawn(ir, authContext)) {
|
||||
for await (const irChunk of streamPlugin.spawn(streamIr, authContext)) {
|
||||
yield irChunk;
|
||||
}
|
||||
} finally {
|
||||
|
||||
@@ -14983,4 +14983,382 @@ describe('Suite 36 — D74 v0.4.1 hotfix regression (maintainer review findings)
|
||||
assert.ok(!/Phase 2 in progress/.test(serverSrc));
|
||||
assert.ok(!/Phase 3 in progress/.test(serverSrc));
|
||||
});
|
||||
|
||||
// ── D75 (v0.4.2) extension — real-machine E2E findings F1, F2, F3, F4, F7 ──
|
||||
// Bugs caught on PI231 + Mac mini 2026-05-26 E2E session. Pre-v0.4.0 reviewers
|
||||
// and the post-v0.4.0 maintainer review all missed these because they reviewed
|
||||
// against spec text, not against real provider CLIs running on a remote machine.
|
||||
|
||||
it('36i (F1) — codex readAuthArtifact() recognises v0.133.0 nested tokens.access_token', () => {
|
||||
// Real codex CLI v0.133.0 auth.json shape: { auth_mode, OPENAI_API_KEY,
|
||||
// tokens: { id_token, access_token, refresh_token, account_id }, last_refresh }
|
||||
// Pre-D75 readAuthArtifact read only top-level access_token → returned null →
|
||||
// OLP falsely reported "auth artifact missing" for fully-logged-in users.
|
||||
const tmpDir = _mkdtempSyncS36(_joinS36(_tmpdirS36(), 'olp-d75-f1-nested-'));
|
||||
const authPath = _joinS36(tmpDir, 'auth.json');
|
||||
_writeFileSyncS36(authPath, JSON.stringify({
|
||||
auth_mode: 'chatgpt',
|
||||
OPENAI_API_KEY: null,
|
||||
tokens: {
|
||||
id_token: 'header.body.sig',
|
||||
access_token: 'eyJ-real-nested-access-token-d75-36i',
|
||||
refresh_token: 'refresh-opaque',
|
||||
account_id: '00000000-0000-0000-0000-000000000000',
|
||||
},
|
||||
last_refresh: '2026-05-26T00:00:00Z',
|
||||
}));
|
||||
const savedPath = process.env.OPENAI_CODEX_AUTH_PATH;
|
||||
process.env.OPENAI_CODEX_AUTH_PATH = authPath;
|
||||
try {
|
||||
const result = codexReadAuthArtifact();
|
||||
assert.ok(result, 'must return non-null for v0.133 nested-shape auth.json');
|
||||
assert.equal(result.accessToken, 'eyJ-real-nested-access-token-d75-36i',
|
||||
'must extract token from tokens.access_token (v0.133.0 nested shape)');
|
||||
} finally {
|
||||
if (savedPath !== undefined) process.env.OPENAI_CODEX_AUTH_PATH = savedPath;
|
||||
else delete process.env.OPENAI_CODEX_AUTH_PATH;
|
||||
_rmSyncS36(tmpDir, { recursive: true, force: true });
|
||||
}
|
||||
});
|
||||
|
||||
it('36j (F1) — codex readAuthArtifact() still reads legacy top-level access_token', () => {
|
||||
// Backwards compat: older codex CLI versions (pre-v0.133) wrote the token
|
||||
// at top level. D75 fix preserves that fallback so upgrades don't regress.
|
||||
const tmpDir = _mkdtempSyncS36(_joinS36(_tmpdirS36(), 'olp-d75-f1-legacy-'));
|
||||
const authPath = _joinS36(tmpDir, 'auth.json');
|
||||
_writeFileSyncS36(authPath, JSON.stringify({
|
||||
access_token: 'legacy-top-level-token-d75-36j',
|
||||
}));
|
||||
const savedPath = process.env.OPENAI_CODEX_AUTH_PATH;
|
||||
process.env.OPENAI_CODEX_AUTH_PATH = authPath;
|
||||
try {
|
||||
const result = codexReadAuthArtifact();
|
||||
assert.ok(result, 'must return non-null for legacy top-level shape');
|
||||
assert.equal(result.accessToken, 'legacy-top-level-token-d75-36j',
|
||||
'must fall back to top-level access_token when tokens.access_token absent');
|
||||
} finally {
|
||||
if (savedPath !== undefined) process.env.OPENAI_CODEX_AUTH_PATH = savedPath;
|
||||
else delete process.env.OPENAI_CODEX_AUTH_PATH;
|
||||
_rmSyncS36(tmpDir, { recursive: true, force: true });
|
||||
}
|
||||
});
|
||||
|
||||
it('36k (F1) — codex readAuthArtifact() returns null for malformed auth.json', () => {
|
||||
// Malformed JSON / missing token / wrong structure → null (not throw).
|
||||
const tmpDir = _mkdtempSyncS36(_joinS36(_tmpdirS36(), 'olp-d75-f1-malformed-'));
|
||||
const savedPath = process.env.OPENAI_CODEX_AUTH_PATH;
|
||||
try {
|
||||
// Case A: unparseable JSON
|
||||
const badJsonPath = _joinS36(tmpDir, 'bad.json');
|
||||
_writeFileSyncS36(badJsonPath, 'not-json-at-all{{{');
|
||||
process.env.OPENAI_CODEX_AUTH_PATH = badJsonPath;
|
||||
assert.equal(codexReadAuthArtifact(), null, 'unparseable JSON → null');
|
||||
|
||||
// Case B: valid JSON but no token field anywhere
|
||||
const noTokenPath = _joinS36(tmpDir, 'no-token.json');
|
||||
_writeFileSyncS36(noTokenPath, JSON.stringify({ auth_mode: 'chatgpt', tokens: {} }));
|
||||
process.env.OPENAI_CODEX_AUTH_PATH = noTokenPath;
|
||||
assert.equal(codexReadAuthArtifact(), null, 'no token in tokens.access_token or top-level → null');
|
||||
|
||||
// Case C: tokens field present but access_token wrong type
|
||||
const wrongTypePath = _joinS36(tmpDir, 'wrong-type.json');
|
||||
_writeFileSyncS36(wrongTypePath, JSON.stringify({
|
||||
tokens: { access_token: 42 },
|
||||
}));
|
||||
process.env.OPENAI_CODEX_AUTH_PATH = wrongTypePath;
|
||||
assert.equal(codexReadAuthArtifact(), null, 'non-string access_token → null');
|
||||
} finally {
|
||||
if (savedPath !== undefined) process.env.OPENAI_CODEX_AUTH_PATH = savedPath;
|
||||
else delete process.env.OPENAI_CODEX_AUTH_PATH;
|
||||
_rmSyncS36(tmpDir, { recursive: true, force: true });
|
||||
}
|
||||
});
|
||||
|
||||
it('36l (F2) — irToCodex() includes --skip-git-repo-check before --model', () => {
|
||||
// codex CLI v0.133.0 sandboxes non-git-repo CWDs by default with:
|
||||
// "Not inside a trusted directory and --skip-git-repo-check was not specified."
|
||||
// OLP servers typically deploy outside a git repo on the operator host. The
|
||||
// --skip-git-repo-check flag must be in args, before --model, so the spawn
|
||||
// succeeds without depending on the install-time CWD.
|
||||
const { args } = irToCodex({
|
||||
irVersion: IR_VERSION,
|
||||
model: 'gpt-5.5',
|
||||
stream: false,
|
||||
messages: [{ role: 'user', content: 'hello' }],
|
||||
});
|
||||
assert.ok(args.includes('--skip-git-repo-check'),
|
||||
'irToCodex().args must include --skip-git-repo-check (F2 fix)');
|
||||
const skipIdx = args.indexOf('--skip-git-repo-check');
|
||||
const modelIdx = args.indexOf('--model');
|
||||
assert.ok(skipIdx > 0 && modelIdx > 0, 'both flags must be present');
|
||||
assert.ok(skipIdx < modelIdx,
|
||||
'--skip-git-repo-check must come before --model (cleaner readability + matches codex v0.133 docs ordering)');
|
||||
// Confirm exec is still first and --json still adjacent
|
||||
assert.equal(args[0], 'exec');
|
||||
assert.equal(args[1], '--json');
|
||||
assert.equal(args[2], '--skip-git-repo-check');
|
||||
assert.equal(args[3], '--model');
|
||||
assert.equal(args[4], 'gpt-5.5');
|
||||
});
|
||||
|
||||
it('36m (F3) — codexChunkToIR recognises v0.133.0 item.completed agent_message', () => {
|
||||
// Real codex v0.133.0 emits the assistant text as one item.completed event:
|
||||
// {"type":"item.completed","item":{"id":"item_0","type":"agent_message","text":"hello world"}}
|
||||
// Pre-D75 codexChunkToIR returned null for this shape → response body had
|
||||
// content: null because the chunk was silently dropped.
|
||||
const result = codexChunkToIR(JSON.stringify({
|
||||
type: 'item.completed',
|
||||
item: { id: 'item_0', type: 'agent_message', text: 'hello world from codex' },
|
||||
}));
|
||||
assert.deepEqual(result, { type: 'delta', content: 'hello world from codex' });
|
||||
});
|
||||
|
||||
it('36n (F3) — codexChunkToIR recognises v0.133.0 turn.completed → stop', () => {
|
||||
// turn.completed marks the end of a codex turn. Map to IR stop.
|
||||
const result = codexChunkToIR(JSON.stringify({
|
||||
type: 'turn.completed',
|
||||
usage: { input_tokens: 12, output_tokens: 7 },
|
||||
}));
|
||||
assert.deepEqual(result, { type: 'stop', finish_reason: 'stop' });
|
||||
});
|
||||
|
||||
it('36o (F3) — codexChunkToIR recognises v0.133.0 turn.failed → error', () => {
|
||||
// turn.failed: { type: 'turn.failed', error: { message: 'X' } }
|
||||
const result1 = codexChunkToIR(JSON.stringify({
|
||||
type: 'turn.failed',
|
||||
error: { message: 'rate limit exceeded' },
|
||||
}));
|
||||
assert.equal(result1?.type, 'error');
|
||||
assert.equal(result1.error, 'rate limit exceeded');
|
||||
// turn.failed with string error
|
||||
const result2 = codexChunkToIR(JSON.stringify({
|
||||
type: 'turn.failed',
|
||||
error: 'simple string error',
|
||||
}));
|
||||
assert.equal(result2?.type, 'error');
|
||||
assert.equal(result2.error, 'simple string error');
|
||||
// turn.failed with no error payload → falls back to default message
|
||||
const result3 = codexChunkToIR(JSON.stringify({ type: 'turn.failed' }));
|
||||
assert.equal(result3?.type, 'error');
|
||||
assert.match(result3.error, /turn\.failed/);
|
||||
});
|
||||
|
||||
it('36p (F4) — cmdStatus reads body.stats.cache.size (not OCP-era body.cache.entries)', () => {
|
||||
// Server payload (server.mjs handleManagementStatus ~line 2092):
|
||||
// { ok, version, ..., stats: { total_requests, active_requests, cache: { hits, misses, size, inflightCount } } }
|
||||
// Pre-D75 cmdStatus formatter read `body.cache.entries` (OCP-era) → always
|
||||
// undefined → output showed "entries=?". Pin the wire contract by grepping
|
||||
// the CLI source so a future server-side rename can't silently re-break.
|
||||
const cliSrc = _readFileSyncS36(_joinS36(import.meta.dirname ?? process.cwd(), 'bin/olp.mjs'), 'utf8');
|
||||
// Must contain the cmdStatus block that reads body.stats.cache (not body.cache directly)
|
||||
assert.ok(/cmdStatus\b/.test(cliSrc), 'cmdStatus function must exist');
|
||||
// The status block must read body.stats?.cache (the actual server payload nesting)
|
||||
assert.ok(/body\.stats\?\.cache/.test(cliSrc) || /body\.stats\.cache/.test(cliSrc),
|
||||
'cmdStatus must read body.stats.cache for status display');
|
||||
// Specifically, the "entries=" output must derive from c.size (matches CacheStore.stats() shape)
|
||||
// Find the status cache line and verify it uses c.size
|
||||
const statusBlock = cliSrc.match(/cmdStatus[\s\S]+?(?=async function cmd|^\}\s*$)/m);
|
||||
assert.ok(statusBlock, 'cmdStatus block extractable');
|
||||
assert.ok(/c\.size/.test(statusBlock[0]),
|
||||
'cmdStatus cache line must use c.size (matches CacheStore.stats() shape)');
|
||||
// And must NOT use the OCP-era c.entries
|
||||
assert.ok(!/c\.entries/.test(statusBlock[0]),
|
||||
'cmdStatus cache line must NOT read c.entries (OCP-era field that does not exist in CacheStore.stats())');
|
||||
// Also pin the CacheStore.stats() shape contract (re-affirms 36e but for status path)
|
||||
const cs = new CacheStore();
|
||||
const stats = cs.stats();
|
||||
assert.ok('size' in stats, 'CacheStore.stats() exposes size');
|
||||
assert.ok(!('entries' in stats), 'CacheStore.stats() does NOT expose entries');
|
||||
});
|
||||
|
||||
it('36q (F7) — per-hop chain `model` field overrides IR model on fallback hop', async () => {
|
||||
// Pre-D75: executeHopFn(provider, model, irReq) used `model` for cache key
|
||||
// + audit but passed irReq UNCHANGED to provider.spawn() → the chain
|
||||
// [{anthropic, claude-X}, {openai, gpt-5.5}]
|
||||
// would spawn codex with --model claude-X (the user's original request)
|
||||
// instead of gpt-5.5 (the chain's per-hop config). This broke cross-provider
|
||||
// fallback wherever the chain assigned different models per provider.
|
||||
//
|
||||
// Strategy: inject a mock openai provider that captures the IR it receives.
|
||||
// Force the anthropic hop to fail with SPAWN_FAILED so the chain advances
|
||||
// to openai. Assert the mock saw model === hop's configured model.
|
||||
|
||||
const { __setAuthConfig: setAC36q, __resetAuthConfig: resetAC36q,
|
||||
__setFallbackConfig: setFC36q, __resetFallbackConfig: resetFC36q,
|
||||
createOlpServer, loadedProviders } = await import('./server.mjs');
|
||||
|
||||
setAC36q({ allow_anonymous: true });
|
||||
|
||||
// Capture for mock openai provider
|
||||
let capturedIR = null;
|
||||
const mockOpenAI = {
|
||||
name: 'openai',
|
||||
displayName: 'Mock OpenAI for F7',
|
||||
contractVersion: '1.0',
|
||||
models: ['gpt-5.5'],
|
||||
auth: { type: 'oauth', storage: 'file', path: '/tmp/fake', refresh: 'auto' },
|
||||
async * spawn(ir, _authContext) {
|
||||
// Snapshot the model field as received
|
||||
capturedIR = { model: ir.model };
|
||||
yield { type: 'delta', role: 'assistant', content: 'mock-openai-response' };
|
||||
yield { type: 'stop', finish_reason: 'stop' };
|
||||
},
|
||||
estimateCost: () => null,
|
||||
quotaStatus: async () => null,
|
||||
healthCheck: async () => ({ ok: true, latencyMs: 0 }),
|
||||
hints: { requiresTTY: false, concurrentSpawnSafe: true, maxConcurrent: 4, cacheable: true },
|
||||
};
|
||||
|
||||
// Mock anthropic provider that always fails immediately with SPAWN_FAILED
|
||||
const mockAnthropic = {
|
||||
name: 'anthropic',
|
||||
displayName: 'Mock Anthropic for F7',
|
||||
contractVersion: '1.0',
|
||||
models: ['claude-sonnet-4-6'],
|
||||
auth: { type: 'oauth', storage: 'file', path: '/tmp/fake', refresh: 'auto' },
|
||||
async * spawn(_ir, _authContext) {
|
||||
throw new ProviderError('mock anthropic always fails (F7 test)', 'SPAWN_FAILED');
|
||||
// eslint-disable-next-line no-unreachable
|
||||
yield { type: 'stop' };
|
||||
},
|
||||
estimateCost: () => null,
|
||||
quotaStatus: async () => null,
|
||||
healthCheck: async () => ({ ok: true, latencyMs: 0 }),
|
||||
hints: { requiresTTY: false, concurrentSpawnSafe: true, maxConcurrent: 4, cacheable: true },
|
||||
};
|
||||
|
||||
// Save + clear loadedProviders, install the mocks
|
||||
const savedProviders = new Map(loadedProviders);
|
||||
loadedProviders.clear();
|
||||
loadedProviders.set('anthropic', mockAnthropic);
|
||||
loadedProviders.set('openai', mockOpenAI);
|
||||
|
||||
// 2-hop chain with DIFFERENT models per hop — the bug only manifests when
|
||||
// hopModel !== requested model on the hop that actually serves.
|
||||
setFC36q({
|
||||
chains: {
|
||||
'claude-sonnet-4-6': [
|
||||
{ provider: 'anthropic', model: 'claude-sonnet-4-6' },
|
||||
{ provider: 'openai', model: 'gpt-5.5' },
|
||||
],
|
||||
},
|
||||
soft_triggers: {},
|
||||
});
|
||||
|
||||
const srv = createOlpServer();
|
||||
let port = 28400 + Math.floor(Math.random() * 100);
|
||||
await new Promise((resolve, reject) => {
|
||||
srv.listen(port, '127.0.0.1', resolve);
|
||||
srv.once('error', e => {
|
||||
if (e.code === 'EADDRINUSE') { port++; srv.listen(port, '127.0.0.1', resolve); srv.once('error', reject); }
|
||||
else reject(e);
|
||||
});
|
||||
});
|
||||
|
||||
try {
|
||||
const r = await fetch({
|
||||
port,
|
||||
method: 'POST',
|
||||
path: '/v1/chat/completions',
|
||||
body: {
|
||||
model: 'claude-sonnet-4-6',
|
||||
messages: [{ role: 'user', content: 'F7 test' }],
|
||||
},
|
||||
});
|
||||
assert.equal(r.status, 200, `expected 200, got ${r.status}: ${r.body.slice(0, 300)}`);
|
||||
assert.equal(r.headers['x-olp-provider-used'], 'openai',
|
||||
'openai must be the serving hop (anthropic failed)');
|
||||
// The F7 assertion: the openai hop received the chain's configured model,
|
||||
// NOT the user's original request model.
|
||||
assert.ok(capturedIR, 'mock openai must have been invoked');
|
||||
assert.equal(capturedIR.model, 'gpt-5.5',
|
||||
`F7: openai hop must receive its chain-configured model 'gpt-5.5', got '${capturedIR.model}'`);
|
||||
} finally {
|
||||
await new Promise(r => srv.close(r));
|
||||
resetFC36q();
|
||||
// Restore original providers
|
||||
loadedProviders.clear();
|
||||
for (const [k, v] of savedProviders) loadedProviders.set(k, v);
|
||||
resetAC36q();
|
||||
}
|
||||
});
|
||||
|
||||
it('36r (F7) — per-hop model override preserved on streaming single-hop path', async () => {
|
||||
// The streaming path (server.mjs ~line 1390-1434) has its own spawn call site
|
||||
// inside sourceFactory that previously also passed `ir` unchanged. F7 fix
|
||||
// wraps it with the same { ...ir, model: streamModel } substitution. This
|
||||
// test exercises a single-hop streaming request and asserts the mock saw
|
||||
// the chain's configured model.
|
||||
|
||||
const { __setAuthConfig: setAC36r, __resetAuthConfig: resetAC36r,
|
||||
__setFallbackConfig: setFC36r, __resetFallbackConfig: resetFC36r,
|
||||
createOlpServer, loadedProviders } = await import('./server.mjs');
|
||||
|
||||
setAC36r({ allow_anonymous: true });
|
||||
|
||||
let capturedModel = null;
|
||||
const mockOpenAIStream = {
|
||||
name: 'openai',
|
||||
displayName: 'Mock Codex for F7 streaming',
|
||||
contractVersion: '1.0',
|
||||
models: ['gpt-5.5'],
|
||||
auth: { type: 'oauth', storage: 'file', path: '/tmp/fake', refresh: 'auto' },
|
||||
async * spawn(ir, _authContext) {
|
||||
capturedModel = ir.model;
|
||||
yield { type: 'delta', role: 'assistant', content: 'streamed-content' };
|
||||
yield { type: 'stop', finish_reason: 'stop' };
|
||||
},
|
||||
estimateCost: () => null,
|
||||
quotaStatus: async () => null,
|
||||
healthCheck: async () => ({ ok: true, latencyMs: 0 }),
|
||||
hints: { requiresTTY: false, concurrentSpawnSafe: true, maxConcurrent: 4, cacheable: true },
|
||||
};
|
||||
|
||||
const savedProviders = new Map(loadedProviders);
|
||||
loadedProviders.clear();
|
||||
loadedProviders.set('openai', mockOpenAIStream);
|
||||
|
||||
// Single-hop chain with a DIFFERENT model than the request — the bug is
|
||||
// visible only when the user requests one model name and the chain remaps it.
|
||||
setFC36r({
|
||||
chains: {
|
||||
'gpt-aliased': [
|
||||
{ provider: 'openai', model: 'gpt-5.5' },
|
||||
],
|
||||
},
|
||||
soft_triggers: {},
|
||||
});
|
||||
|
||||
const srv = createOlpServer();
|
||||
let port = 28500 + Math.floor(Math.random() * 100);
|
||||
await new Promise((resolve, reject) => {
|
||||
srv.listen(port, '127.0.0.1', resolve);
|
||||
srv.once('error', e => {
|
||||
if (e.code === 'EADDRINUSE') { port++; srv.listen(port, '127.0.0.1', resolve); srv.once('error', reject); }
|
||||
else reject(e);
|
||||
});
|
||||
});
|
||||
|
||||
try {
|
||||
const r = await fetch({
|
||||
port,
|
||||
method: 'POST',
|
||||
path: '/v1/chat/completions',
|
||||
body: {
|
||||
model: 'gpt-aliased',
|
||||
stream: true,
|
||||
messages: [{ role: 'user', content: 'F7 streaming test' }],
|
||||
},
|
||||
});
|
||||
assert.equal(r.status, 200, `expected 200, got ${r.status}: ${r.body.slice(0, 200)}`);
|
||||
assert.equal(capturedModel, 'gpt-5.5',
|
||||
`F7 streaming: openai must receive chain-configured model 'gpt-5.5', got '${capturedModel}'`);
|
||||
} finally {
|
||||
await new Promise(r => srv.close(r));
|
||||
resetFC36r();
|
||||
loadedProviders.clear();
|
||||
for (const [k, v] of savedProviders) loadedProviders.set(k, v);
|
||||
resetAC36r();
|
||||
}
|
||||
});
|
||||
});
|
||||
|
||||
Reference in New Issue
Block a user