* fix(daemon): P1-1 guard proc.stdin against EPIPE crash
In spawnClaudeProcess, attach an error listener on proc.stdin BEFORE
the write/end calls so an EPIPE (child closed stdin mid-write) is
swallowed and logged rather than thrown as an unhandled exception.
The existing proc.on("error") listener is on the ChildProcess object,
NOT on the stdin Writable — it does not catch stdin write errors.
Hardening per OCP code audit; entry-surface contract unchanged for
single-user default path.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(tui): P1-3 remove tool_use from isTerminalLine; only turn_duration is terminal
In interactive TUI mode, stop_reason=tool_use does NOT mean the turn is
complete. Claude handles the tool call internally and continues generating —
the transcript advances to another assistant entry. Treating tool_use as
terminal truncated tool-using turns mid-flight.
Only {type:"system", subtype:"turn_duration"} is the authoritative
completion marker (claude CLI v2.1.157+ interactive session transcript).
Updated two unit tests that previously asserted tool_use → true; they now
assert false (the correct behaviour). The real-fixture terminal detection
test is unaffected because the fixture uses turn_duration.
Hardening per OCP code audit; TUI path behaviour fix.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(tui): P2-5 add -l (literal) flag to send-keys prompt paste
A prompt that equals a tmux key token (e.g. "C-c", "Escape") would be
interpreted as that key binding rather than typed as literal text.
The -l flag forces literal character-by-character input. The separate
Enter key event afterward deliberately omits -l so tmux sends a real
carriage-return keypress to submit the prompt line.
Authority: tmux send-keys(1) § -l flag. Hardening per OCP code audit.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(audit): String-coerce parsed.error + stale tool_use comment (review fast-follow)
Folds the 2 minor findings from the independent review of the audit fixes:
- String(parsed.error) before .slice/message in callClaude + callClaudeStreaming
(defensive: claude could emit a non-string result/error_message).
- correct the readTuiTranscript comment that still listed tool_use as terminal.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Reader + driver + home modes + server wiring + entrypoint hardening. 5 independent
reviews folded; e2e green; default stream-json path byte-identical when flag off.
See PR description + ADR 0007 for the full layer/authority/security detail.
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* wip(rescue): OCP Phase 6c stream-json port + opus 4.8 (uncommitted /tmp work)
Rescue commit — preserves the subagent-produced Phase 6c port (claude -p →
stream-json + --system-prompt + NDJSON parser) and opus 4.8 model addition
done 2026-05-30 in a /tmp clone (volatile). NOT reviewed-for-merge; this is
preservation only on a WIP branch so a /tmp reboot does not lose the work.
Strategy context: OCP is being made the first-mover for TUI-mode (users are
on OCP). Phase 6c here is the DEFAULT-path foundation (stream-json = sdk-cli =
the "Max-user stable mode" fallback). TUI-mode (cc_entrypoint=cli subscription
bridge) lands on top as an opt-in. Re-review before merging to main.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(phase6c): fold review findings — scrub hostname, README prose, parser tests, version citation
P3-1 (binding): scrub private hostname "PI231" from server.mjs comment; replace
with role-based term "the test server". Repo is public — no machine-specific
hostnames may appear.
P3-2: update README.md "How It Works" diagram and prose from stale "claude -p"
to "claude --output-format stream-json", matching the Phase 6c spawn change already
in server.mjs and CHANGELOG.
P2-2: update all five authority citations in server.mjs from bare "claude CLI v2.1.104"
to "claude CLI § <flag> (ported from OLP, verified v2.1.104; behavior stable through
v2.1.158)" — honest about OLP provenance and version range, without inventing new facts.
P2-1: add "Stream-JSON parsers" suite to test-features.mjs (17 new tests, 84 → 101
total). Copies parseStreamJsonLines and parseStreamJsonEvent verbatim from server.mjs
with a MIRRORS header comment; logEvent stubbed to avoid live-server side-effects.
Covers: content_block_delta deltas, assistant-aggregate fallback, no-double-count guard,
aggregate-only short responses, partial-line buffering across two chunks, is_error result
variants, malformed/non-JSON lines, system/user/unknown event types.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Closes the structural side of the port-drift cascade addressed by
v3.16.2/v3.16.3. Those releases reverted the literal line-by-line; this
one removes the invitation to drift.
Changes:
* NEW lib/constants.mjs — exports DEFAULT_PORT=3456, LOCAL_HOST,
OPENAI_API_BASE, LOCAL_PROXY_URL.
* server.mjs / setup.mjs / scripts/upgrade.mjs / scripts/doctor.mjs
(x2) / scripts/sync-openclaw.mjs all import DEFAULT_PORT from
lib/constants.mjs instead of hardcoding "3456".
* .github/workflows/alignment.yml:
- path filter extended to setup.mjs, scripts/**, lib/**,
ocp, ocp-connect.
- NEW job port-spot hard-fails any PR that introduces a hardcoded
"3478" or "3456" literal outside EXEMPT_REGEX (lib/constants.mjs,
test-features.mjs, ocp/ocp-connect bash CLIs, docs, the workflow
itself).
* Doc-comment rewording so CI grep finds zero hits.
No behavior change for any user. CLAUDE_PROXY_PORT env var still wins
at runtime; only the unset-env fallback now flows through one constant.
ALIGNMENT.md note: server.mjs change is one import + one literal swap,
mechanical. No cli.js operation changed; the citation requirement does
not apply.
cli.js: not applicable — mechanical refactor, no behavior change.
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Identified during privacy/security audit (.claude/research/ocp-security-audit.md
§3 "/api/usage 全量泄露所有 key 时序"). Previously, any caller authenticated as
admin received the full per-key usage byKey + recent + timeline block, which
means a brief admin-token compromise exposes every household member's request
volume, timing, and which-model — metadata, but sensitive metadata.
This is a deliberate breaking change for least-privilege.
Behavior matrix (post-change):
Caller | Default scope | ?all=true
----------------------------------------|-------------------|---------------------
anonymous (PROXY_ANONYMOUS_KEY) | own ("anonymous") | ignored (still own)
authenticated non-admin key | own (key.name) | ignored (still own)
admin (no flag) | own ("admin") | n/a
admin with ?all=true | n/a | full byKey/recent
localhost-no-token / "local" | own ("local") | full (isAdmin=true)
Response shape is unchanged except for an additional advisory `scope` object
(`{self, all}`); existing fields keep their structure so dashboards and scripts
continue to parse cleanly — they just see less data unless they opt in.
Audit-friendly addition: when admin opts into `?all=true`, the server now
emits `logEvent("info", "admin_usage_full_scope", { caller, ip })` so every
privilege-escalation moment is recorded.
Backward-compat warning for admin scripts:
existing cron jobs / scripts that call `/api/usage` to inventory all keys must
add `?all=true` after this change. Default-self is the new safe behavior.
Dashboard (dashboard.html): adds a "Show all keys" checkbox in the Usage by
Key section. Hidden by default; revealed only when refreshKeys() succeeds
(same admin gate as the keys-management section). Toggle state persists in
localStorage as `ocp_usage_show_all`.
cli.js citation
---------------
This is OCP-internal admin API behavior — there is no `cli.js` operation to
cite. ALIGNMENT.md Rule 2 (the rule that limits OCP to operations that exist
in `cli.js`) does not constrain access-control rules for the OCP server's
own admin endpoints.
Smoke test (isolated test server on :3489, prod :3478 untouched):
- node --check server.mjs SYNTAX_OK
- npm test 43/43 passed
- alignment.yml blacklist grep BLACKLIST_CLEAR
- LAN scope matrix alice/bob own only; alice ?all=true denied;
anon ?all=true denied; admin all=true full +
audit log emitted
Iron Rule 10
------------
Author cannot self-approve. A separate fresh-context reviewer must read the
diff and confirm the cli.js-not-applicable statement is correct before merge.
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Credential-bearing OCP artifacts were created at default umask (0644/0755),
making them world-readable on multi-user hosts. This commit hardens them at
writer time and adds an idempotent startup reconciliation so existing prod
installs are fixed automatically on next service restart.
Changes:
- keys.mjs: mkdirSync(OCP_DIR, { mode: 0o700 }) + chmodSync after to handle
pre-existing dirs; chmodSync(DB_PATH, 0o600) after first getDb() open.
- setup.mjs: chmodSync(plistPath, 0o600) after writeFileSync on macOS;
chmodSync(unitPath, 0o600) after writeFileSync on Linux.
- server.mjs: _tightenFileModesIfPossible() reconciliation block — idempotently
chmods ~/.ocp (700), ~/.ocp/admin-key (600), ~/.ocp/ocp.db (600) on startup;
emits a single info-level log line when any file is tightened; ignores ENOENT
and wraps EPERM in a warn log so startup is never crashed by chmod failure.
Backward compat: all files remain accessible to the same-user owner; 0o600
is still fully readable/writable by the process. Existing prod boxes with
old 0644 ~/.ocp directories get fixed-up on next launchd/systemd restart
without any manual intervention.
cli.js citation: this change is OCP-internal file-permission hardening only.
No cli.js function corresponds to chmod or credential-file management. This
is an OCP-local security improvement that is out of scope for the cli.js
citation requirement per ALIGNMENT.md Rule 2; the cli.js boundary applies
to proxy protocol and API surface changes, not host-filesystem hardening.
Identified during privacy/security audit — .claude/research/ocp-security-audit.md.
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
**Bug**: the sessions Map used the raw client-supplied conversationId string as
its key. Two callers with different API keys (or one anonymous + one
authenticated) using the same session_id="default" collided in the Map,
sharing a cli.js subprocess and conversation history — a cross-tenant context leak.
**Fix**: introduce _sessionKey(conversationId, keyName) → "${keyName}|${conversationId}".
Replace every sessions Map site (has / set / get / delete) with the namespaced key.
keyName comes from req._authKeyName (set by auth middleware). Anonymous callers
produce "anon|<id>"; admin produces "admin|<id>"; per-key callers produce
"<keyName>|<id>" — matching the convention used by cacheHash() for per-key cache
isolation (D1, v3.13.0).
**cli.js citation — not applicable**: This PR changes OCP-internal session lifecycle
bookkeeping (the sessions Map that OCP maintains to thread --resume flags across
requests). There is no corresponding cli.js operation to cite. ALIGNMENT.md Rule 2
(limiting OCP to operations cli.js performs) does not constrain in-process state
representation. Same pattern as #75.
**Scope of changes (server.mjs only)**:
- New helper: _sessionKey() — 3 lines after sessions Map declaration
- spawnClaudeProcess(): accepts keyName param; all 4 sessions Map calls → _sessionKey
- handleSessionFailure(): uses sessionKey (not raw conversationId) for sessions.delete
- callClaude(): accepts and forwards keyName to spawnClaudeProcess
- callClaudeStreaming(): forwards authInfo.keyName to spawnClaudeProcess
- Request handler: both callClaude() call sites pass req._authKeyName
- Cleanup interval + GET /health + GET /sessions: strip "keyName|" prefix before log/display
**Smoke test**:
- node --check server.mjs → SYNTAX OK
- npm test → 43/43 passed, 0 failed
- Hand-traced has/set/get/delete paths: cross-key collision absent ✓
Identified during privacy/security positioning brainstorm — `.claude/research/ocp-security-audit.md`
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Real-world macOS dev machines using nvm-managed Node hit a startup FATAL
because the hardcoded candidate list in resolveClaude() only covered
homebrew, /usr/local, /usr/bin, and ~/.local/bin. With Claude CLI
installed at $HOME/.nvm/versions/node/<v>/bin/claude, the launchd job
failed without manual CLAUDE_BIN injection.
Fix: extend the candidate list with user-local Node version manager
paths — nvm (with default-alias), fnm, asdf, and npm-prefix-relocated
$HOME/.npm-global/bin. The existing CLAUDE_BIN env override and `which`
fallback are preserved; resolution order is now explicit CLAUDE_BIN >
hardcoded list > nvm/fnm/asdf > which > FATAL (with the message
upgraded to mention CLAUDE_BIN as a hint).
This is OCP-internal binary discovery — there is no `cli.js` operation
to cite. ALIGNMENT.md Rule 2 (the rule that limits OCP to operations
that exist in `cli.js`) does not constrain runtime path discovery for
the OCP server itself.
Smoke tests:
- default (no CLAUDE_BIN): picks /opt/homebrew/bin/claude (unchanged)
- CLAUDE_BIN=/nonexistent/claude: fail-fast preserved
- HOME=/tmp/fakenvm with synthetic .nvm tree: candidate list contains
the fake nvm path; alias-default unshift logic verified
- npm test: 43/43 unit tests pass
- node --check server.mjs: OK
- alignment.yml blacklist grep: no hits
Identified during fresh-state Round 2 testing on MacBook.
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
cli.js does not perform proxy-layer stampede protection. The singleflight
layer is a value-add proxy operation that exists only inside OCP, between
concurrent client requests and the single upstream cli.js spawn. It does
not introduce, alter, or remove any endpoint, header, request field, or
response field that cli.js emits or expects — no client-observable wire
shape change.
Justification under ALIGNMENT.md Rule 2: the singleflight Map deduplicates
concurrent identical non-streaming cache-miss requests so only one cli.js
spawn runs per unique hash window. All followers receive the same resolved
(or rejected) content. This is a proxy-internal concurrency optimization,
not a wire-level protocol change.
Spec: docs/superpowers/specs/2026-05-07-cache-upgrade-design.md (D4)
Changes:
- keys.mjs: add singleflight(hash, fn) + getInflightStats() exports;
in-memory Map cleared via Promise.finally() on each settlement
- server.mjs: import singleflight + getInflightStats; wrap non-streaming
cache-enabled path through singleflight with inner recheck; add TODO
comment in callClaudeStreaming (streaming-path dedup is explicitly out
of scope for v3.13.0, see spec D4 streaming caveat); extend /cache/stats
to return inflight + requesters fields (additive, no removed fields)
- test-features.mjs: 7 new PR-B singleflight tests covering basic dedup,
failure fan-out, map cleanup (success + failure), different-hash
independence, getInflightStats shape, and sequential-call non-sharing;
all 31 tests pass (24 existing + 7 new)
Streaming-path singleflight is explicitly out of scope; TODO left in
callClaudeStreaming for a future follow-up ticket.
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
* docs(governance): ADR 0005 (no multi-provider) + cache upgrade spec
Governance prelude for the cache upgrade work:
- docs/adr/0005-no-multi-provider.md — locks in the decision that
OCP stays single-provider (Anthropic via cli.js spawn). Cache
improvements are explicitly in scope (decision §3); multi-provider
refactor is explicitly out of scope, with three documented trigger
conditions for revisiting.
- docs/adr/README.md — index updated to reference 0005.
- docs/superpowers/specs/2026-05-07-cache-upgrade-design.md — design
for the cache upgrade work split into PR-A (per-key isolation,
cache_control bypass, chunked stream replay) and PR-B (singleflight
stampede protection). Each design decision has a written rationale.
server.mjs is not modified; this commit is doc-only and therefore
exempt from the cli.js citation requirement (ALIGNMENT.md Rule 5
applies only to commits that touch server.mjs).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* feat(cache): per-key isolation, cache_control bypass, chunked stream replay
cli.js does not perform response caching at the proxy layer. OCP's response
cache is a value-add operation internal to OCP, between the wire and the
cli.js spawn. It does not introduce, rename, or alter any endpoint, header,
request field, or response field that cli.js emits or expects — this change
qualifies under ALIGNMENT.md Rule 2's value-add carve-out for non-wire-
affecting proxy operations. no client-observable wire shape change.
Spec: docs/superpowers/specs/2026-05-07-cache-upgrade-design.md
D1 — Per-key cache isolation (cacheHash v2 format)
keys.mjs: prepend `v2|k:<keyId or "anon">|` before existing hash fields.
Backward-compatible: absent/null/empty keyId folds to "anon".
v1-format rows in response_cache are abandoned naturally; TTL cleanup at
server.mjs:185 reaps them within one window. No migration needed.
server.mjs: pass keyId: req._authKeyId at the single cacheHash call site
(line ~1221).
D2 — cache_control bypass
keys.mjs: export hasCacheControl(messages) — walks messages and nested
content arrays for presence of cache_control field.
server.mjs: if hasCacheControl(messages) is true, set req._cacheHash = null
and log cache_skipped{reason: cache_control_present}; existing
`if (CACHE_TTL > 0 && req._cacheHash)` guards on write-back handle the skip.
D3 — Chunked stream replay (80 codepoints/chunk, no artificial delay)
server.mjs: replace single-chunk cached.response emission with an
Array.from(cached.response) loop in steps of CACHE_REPLAY_CHUNK_SIZE=80.
Array.from ensures multibyte UTF-8 codepoints (e.g. CJK) are never split.
Tests: 12 new cases in test-features.mjs (36 total, 0 failed).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Adds structured logging for the two forensic candidates so the eventual
fixes can be evidence-driven (per #41/#42 bodies: "Evidence needed before
code change — do not patch speculatively"). NO bug fix in this PR — the
stale-create-entry behavior (#41) and the lastUsed-resistant-expiry
behavior (#42) are intentionally left unchanged so they can be observed
in production /logs.
Three changes (server.mjs):
1. Session-create branch records `firstSeen` alongside `lastUsed` (same
timestamp), giving the sweep loop an absolute-age signal independent
of the actively-bumped `lastUsed` field. Resume branch is untouched
so `firstSeen` is preserved across resumes.
2. handleSessionFailure now emits a structured `session_failure` event
with `mode: "resume"` (action: "deleted") OR `mode: "create"`
(action: "kept"). The "kept" branch makes #41's "stale create entries
never deleted" pattern visible in /logs without changing behavior.
3. TTL sweep emits `session_expired` (info, with idleMs + ageMs) on
actual expiry, and `session_long_lived` (warn) when ageMs > 4×TTL
but the entry is not being expired this tick. The latter is the
#42 evidence trigger — a session whose lastUsed keeps getting bumped
and so resists idle-based expiry.
cli.js citation: N/A — logEvent payloads are OCP-internal observability,
never sent on the wire to the Anthropic API. ALIGNMENT.md Rule 2 (no
invention) is scoped to endpoints/headers/request/response fields, not
to internal server logging.
Independent reviewer: fresh-context sonnet APPROVE_WITH_MINOR.
Two minor notes — repeated `session_long_lived` emissions per 60s
sweep tick is expected (downstream analysis dedupes by conversationId).
LOC: +19 / -3 = +16 net.
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
* docs(spec): design for #47 SSE heartbeat on streaming path
Draft spec for an opt-in idle-watchdog SSE heartbeat covering both
pre-first-byte and mid-stream silent windows. Default disabled,
controlled by CLAUDE_HEARTBEAT_INTERVAL. Targets ~40 LOC.
Decisions captured: D1 whole-stream reset-on-byte; D2 SSE comment
frame; D3 default disabled; D4 relocate ensureHeaders() to post-spawn;
D5 X-Accel-Buffering: no on both SSE header sites; D6 single log line
per affected request.
Scope-locked: does not touch CLAUDE_TIMEOUT semantics, the separate
server.mjs:480-489 dangling-client bug, issues #41/#42, or the
non-streaming path.
Includes privacy preflight and cloud-testing plan per maintainer
feedback.
Refs: #47
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* docs(plan): implementation plan for #47 SSE heartbeat
6 phases: pre-work (file 480-bug), implementation (5 tasks on
feat/47-sse-heartbeat), opus fresh-context review, cloud verification
on Mac rig, privacy preflight, PR+release.
Each implementation task carries concrete code, syntax check, and a
scoped commit message. LOC budget enforced in Task 1.6 gate (~45
server.mjs lines max). Reviewer checklist scopes scope-lock, ALIGNMENT,
privacy, and heartbeat-cannot-abort discipline.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* feat(server): add startHeartbeat helper + HEARTBEAT_INTERVAL env var
Per design doc (refs #47). Helper is a per-request idle watchdog that
emits `: keepalive\n\n` SSE comment frames; returns a {reset, stop} handle.
No wiring yet — helper is unused, safe to commit in isolation.
cli.js citation: N/A — SSE response shaping is an OCP-owned translation
layer, not a cli.js operation. See AGENTS.md and ALIGNMENT.md Rule 2.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* refactor(server): eagerly send SSE headers post-spawn (D4, refs #47)
Moves the ensureHeaders() call from "on first stdout byte" to
"immediately after successful spawn." This is a prerequisite for the
heartbeat covering the pre-first-byte silent window (the 'processing
large contexts' failure mode in #47).
Behavioral consequence: the narrow "spawn succeeded but subprocess died
before any byte" branch at server.mjs:610-611 becomes effectively dead
in the common case. The post-headers SSE-stop path (612-619) handles
it instead. The branch remains defensively for the client-closed-before-
ensureHeaders race.
Isolated commit per design doc §D4 so reviewer can focus on this one
behavior change.
cli.js citation: N/A — SSE header emission is OCP response-shaping.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* feat(server): wire heartbeat into streaming path (refs #47)
- sendSSE() accepts optional hb handle and calls hb.reset() before write
- callClaudeStreaming starts heartbeat after ensureHeaders() and passes
hb to the three streaming sendSSE call sites
- All three exit paths (proc close, proc error, res close) call
hb.stop() to guarantee timer cleanup; no-op handle when disabled
means zero runtime cost when CLAUDE_HEARTBEAT_INTERVAL=0
Heartbeat never aborts — only writes comment frames and re-arms. Aligns
with v3.3 timeout discipline (single CLAUDE_TIMEOUT, no secondary
client-killing timers).
cli.js citation: N/A — SSE response shaping is OCP translation layer.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* feat(server): add X-Accel-Buffering: no to SSE response headers (refs #47)
nginx (and many LBs / Cloudflare) default to proxy_buffering=on, which
would buffer heartbeat comment frames indefinitely and defeat the
feature silently. This header hints no-buffering; other stacks ignore
it. Applied at both SSE header sites (real streaming + cache-hit).
cli.js citation: N/A — response header shaping is OCP translation layer.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* chore(release): v3.12.0 — streaming heartbeat (refs #47)
Bundles the release-kit companion files per Iron Rule 5.2 / 11 example:
version bump across package.json + ocp-plugin + openclaw.plugin.json,
CHANGELOG v3.12.0 section, README env var row + "Streaming heartbeat"
explainer.
Tag push to v3.12.0 triggers .github/workflows/release.yml to create
the GitHub Release automatically.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(server): ensureHeaders returns true when already sent (refs #47)
Phase 3 smoke test revealed every content chunk was being dropped
after the D4 eager ensureHeaders() call: the stdout.on('data') guard
`if (!ensureHeaders()) return;` early-returned on every chunk because
ensureHeaders returned false for the already-sent case (conflated with
the dead-connection case).
Split the two conditions explicitly: return false only when res is
ended/destroyed; return true when headers are (already or now) sent.
This also fixes a latent multi-chunk bug on main that was masked
because claude CLI typically outputs in one stdout chunk.
Verified: node -c server.mjs; subsequent re-run of Phase 3 smoke test
now sees streaming content chunks + heartbeat frames.
cli.js citation: N/A — SSE response shaping is OCP translation layer.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Per forensic analysis in #37, the timeout handler at server.mjs:471
called `proc.kill('SIGTERM')` without decrementing the concurrency
counter (`stats.activeRequests`). If the subprocess was stuck in a
syscall (e.g. MCP I/O) and ignored SIGTERM, its slot was not freed
until — and only if — the `proc.on('close')` handler eventually fired
after the 5s SIGKILL escalation successfully reaped the child.
Real-world observation (#37) shows 3 stuck claude subprocesses running
20 minutes to 2h 45m, exhausting the 8-slot pool and returning
`500 concurrency limit reached (8/8)` on every subsequent request.
Fix: attach `cleanup` to `proc.once('exit', ...)`. The 'exit' event
fires before 'close' and runs on every termination path — normal
completion, error, SIGTERM, SIGKILL, crash — even if stdio pipes
stay open. `cleanup()` is already idempotent (guarded by `cleaned`
flag), so this listener is safe alongside the existing 'close' and
'error' paths that also call it.
ALIGNMENT: cli.js has no equivalent subprocess-pool / concurrency-
limit logic — cli.js is a single-user CLI, the process being pooled,
not the pool itself. Per ALIGNMENT.md Rule 2, this fix is proxy-
internal (subprocess lifecycle management) with no cli.js equivalent
to cite; documented here instead. No wire-protocol, HTTP surface, or
response shape change (Rule 3 preserved).
Release kit: bumps version to 3.11.1 and adds CHANGELOG entry so the
auto-release workflow tags v3.11.1 on merge. README version refs are
feature-gate markers for v3.11.0 (models.json SPOT / auto-sync) and
remain historically accurate — no README change needed.
Fixes#37.
Co-authored-by: dtzp555 <dtzp555@gmail.com>
Adds scripts/sync-openclaw.mjs which reconciles
config.models.providers["claude-local"].models and
config.agents.defaults.models["claude-local/*"] with models.json.
Hooked into `ocp update` between git pull and proxy restart, non-fatal
(sync failure does not abort the update; the gateway still restarts and
/v1/models still works).
Scope boundaries (honoring the no-OpenClaw-source-edit constraint):
- Only touches claude-local provider block and claude-local/*
alias keys. baseUrl / api / authHeader preserved on existing
installs (user may have customized port). All other providers
and top-level config keys untouched.
- Creates a timestamped backup (openclaw.json.bak.<ms>) before every
write. No-op path (already in sync) skips backup.
- Exits 0 with a skip message when ~/.openclaw/openclaw.json does not
exist (non-OpenClaw users).
server.mjs change: a 15-line passive self-check added inside the
existing server.listen() callback. It only reads the OpenClaw config
and emits console.warn on drift. No network/endpoint surface added, no
new headers, no new routes, no new Claude-CLI-call path. Per
ALIGNMENT.md Rule 2 this is not an operation cli.js performs, so no
cli.js:NNNN citation is required. The added `existsSync` import from
node:fs is already part of the node:fs module surface.
Independent reviewer: Tao pre-approved the 3-PR plan; Iron Rule 10
waived for this task-scoped execution.
Co-authored-by: Tao Deng <dtzp555@gmail.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Pure refactor. No API surface change. MODEL_MAP and MODELS in server.mjs
are now derived from models.json; /v1/models response is byte-equivalent
to the previous hardcoded table (verified via equality check on the
resulting Object.entries sorted).
setup.mjs MODEL_ID_MAP / MODELS / MODEL_ALIASES also derived from
models.json, fixing a latent drift where setup.mjs still listed
claude-opus-4-6 / claude-haiku-4 (v3.0-era) while server.mjs had
already moved to opus-4-7 + haiku-4-5-20251001 in v3.10.0.
server.mjs change scope: no network/endpoint surface modified. No new
env vars, headers, or routes. This is a pure data-source refactor of
two existing top-level constants (MODEL_MAP, MODELS). No cli.js
operation is being added or changed, so per ALIGNMENT.md Rule 2 this
requires no cli.js:NNNN citation — the change does not touch the
Claude-CLI-call boundary.
Independent reviewer: Tao pre-approved the 3-PR plan that contains this
refactor; review is waived for this PR under that pre-approval
(CLAUDE.md Iron Rule 10 exception, task-scoped).
Unblocks PR B (scripts/sync-openclaw.mjs), which needs a single source
of truth to reconcile with on `ocp update`.
Co-authored-by: Tao Deng <dtzp555@gmail.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Adds claude-opus-4-7 as an available model and promotes the short
aliases 'opus' and 'claude-opus-4' to point at 4.7 (following the
same 'latest under the short alias' pattern as sonnet/haiku). Explicit
claude-opus-4-6 remains available for pinned usage.
Claude Code alignment evidence (binary-era; see note):
- Anthropic /v1/models (2026-04-20): returns id='claude-opus-4-7',
display_name='Claude Opus 4.7', 1M input / 128K output,
created_at=2026-04-14.
- Claude Code v2.1.114 /home/opc/.npm-global/lib/node_modules/
@anthropic-ai/claude-code/bin/claude.exe strings table contains
'claude-opus-4-7'.
- Live probe: 'claude -p --model claude-opus-4-7' returns a valid
response (verified 2026-04-20).
Note on ALIGNMENT.md grep rule: Claude Code 2.1.90+ ships a native
binary (claude.exe) instead of cli.js JavaScript. The grep-based
verification in ALIGNMENT.md Rule 1 is substituted here with
(a) binary 'strings' extraction and (b) Anthropic /v1/models as
an independent authoritative source. A follow-up constitution
amendment will codify the binary-era verification procedure.
Co-authored-by: Oracle Public Cloud User <opc@instance-20230820-1333.subnet07301351.vcn07301351.oraclevcn.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Follow-up to #21. The restored header-based fetchUsageFromApi() copied
the golden 47e39d7 implementation verbatim, which used x-api-key auth
because v3.0.0 targeted API-key users (sk-ant-api03-*). Current OCP
uses OAuth tokens (sk-ant-oat01-*) from Claude Pro/Max subscriptions,
so x-api-key with an OAuth token returns 401, breaking /usage.
Claude Code cli.js alignment evidence:
- For OAuth calls, headers are:
Authorization: Bearer <accessToken>
anthropic-beta: oauth-2025-04-20
- Constant qJ="oauth-2025-04-20" in cli.js
- Multiple call sites confirmed (e.g. /v1/files upload, /v1/sessions)
Fix: replace x-api-key with Authorization: Bearer + add anthropic-beta
header. Matches the exact headers cli.js sends for every OAuth request.
ALIGNMENT.md compliance: change aligns OCP with cli.js; CI blacklist
unaffected.
Co-authored-by: Oracle Public Cloud User <opc@instance-20230820-1333.subnet07301351.vcn07301351.oraclevcn.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Removes the stale-cache fallback branches in handleUsage() and
handleStatus() originally introduced by cb6c2a8 (2026-04-12).
Background: cb6c2a8 was a compensation for b87992f's hallucinated
/api/oauth/usage endpoint starting to 429 within 24h of deployment.
Since PR B restores the correct header-based endpoint from /v1/messages,
this compensation is now dead code masking a problem that no longer exists.
Changes:
- handleUsage: remove `else if (usageCache.data)` stale fallback
- handleStatus: remove `else if (usageCache.data)` stale fallback
- USAGE_CACHE_TTL: already reset to 5min in PR B's rewritten block
(cb6c2a8 had bumped it to 15min as further compensation)
Depends on #21 (PR B: restore header-based /usage). Merge PR B first.
Co-authored-by: Oracle Public Cloud User <opc@instance-20230820-1333.subnet07301351.vcn07301351.oraclevcn.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Replaces fetchUsageFromApi() hallucinated /api/oauth/usage endpoint
with the original header-based approach: POST /v1/messages with
max_tokens=1 and extract anthropic-ratelimit-unified-{5h,7d}-{utilization,reset}
from response headers.
Drift: b87992f (2026-04-11) introduced /api/oauth/usage — an endpoint
that does not exist in Claude Code cli.js. Hallucinated. Within 24h
it started returning 429, which cb6c2a8 attempted to mask with stale
cache + extended TTL (cleaned up in follow-up PR C).
Golden reference: 47e39d7 v3.0.0 (2026-03-24) — correct implementation
using anthropic-ratelimit-unified-* headers.
Claude Code cli.js alignment evidence:
- function vE4 iterates [["five_hour","5h"],["seven_day","7d"]]
- reads headers anthropic-ratelimit-unified-${key}-utilization
and anthropic-ratelimit-unified-${key}-reset
- cli.js path: /home/opc/.npm-global/lib/node_modules/@anthropic-ai/claude-code/cli.js
Preserved modernisations from post-47e39d7 work:
- getOAuthCredentials(): keychain + Linux ~/.claude/.credentials.json
- NEW: CLAUDE_CODE_OAUTH_TOKEN env var fallback (highest precedence)
- OAuth refresh on 401 / pre-emptive on expiry
- NEW: exponential refresh backoff 60s -> 3600s to prevent tight
retry loops (previously burned rate-limit in seconds after 401)
Added ALIGNMENT anchor comment at top of block referencing cli.js vE4
and ALIGNMENT.md, so future refactors can grep before re-drifting.
Co-authored-by: Oracle Public Cloud User <opc@instance-20230820-1333.subnet07301351.vcn07301351.oraclevcn.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Add two new features for LAN sharing governance:
Quota (budget control):
- Per-key daily/weekly/monthly request limits (NULL = unlimited)
- Idempotent schema migration for quota columns
- Single-query check (SUM/CASE) for all 3 periods — no N+1
- PATCH /api/keys/:id/quota (partial update, input validation)
- GET /api/keys/:id/quota (current limits + usage)
- 429 with structured error when exceeded
- Only applies to identified per-key users, not admin/anonymous
Response cache:
- SHA-256 hash of model + messages + temperature/max_tokens/top_p
- Opt-in via CLAUDE_CACHE_TTL env var (0 = disabled, default)
- Cache hit serves both streaming (simulated SSE) and non-streaming
- Streaming responses accumulated and cached on success
- Skips multi-turn sessions (conversationId present)
- GET /cache/stats, DELETE /cache admin endpoints
- Runtime-tunable cacheTTL via PATCH /settings
- 10-minute periodic cleanup of expired entries
Bug fix discovered during testing:
- SQLite datetime('now') stores 'YYYY-MM-DD HH:MM:SS' but JS
.toISOString() produces 'YYYY-MM-DDTHH:MM:SS.sssZ'. String
comparison breaks for same-day ranges. Added sqliteDatetime()
helper for correct format matching.
Code review fixes:
- DELETE /api/keys/:id no longer shadows /quota sub-routes
- updateKeyQuota uses partial UPDATE (only SET provided fields)
- cacheHash includes temperature/max_tokens/top_p in hash
- Replaced raw getDb() in server.mjs with findKey() encapsulation
- Unified UTC midnight calculation across checkQuota/getKeyQuota
Includes 24-test integration suite (test-features.mjs).
Co-authored-by: Tao Deng <dtzp555@gmail.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Implements issue #12 section 14 Path A. Lets OCP admin designate a
single well-known "anonymous" key that bypasses validateKey() while
keeping AUTH_MODE=multi. Gives OpenClaw multi-agent clients (which
MUST send a non-empty Authorization header per their per-agent
auth-profiles schema) a way to connect without every user needing
a personal key.
## Background
After PR-1 (issue #12), we know OpenClaw multi-agent setups
require a per-agent auth-profiles.json with a non-empty `key`
field. OCP multi-mode rejects any non-empty Bearer token that
isn't in the keys database (server.mjs line 1152), which creates
a deadlock: the only way for OpenClaw to use OCP is with a real
admin-issued key.
Path A resolves this by letting the admin opt in to a public
anonymous key that clients can auto-discover via /health. OpenClaw
writes this key into its agent profiles just like a real key; OCP
server short-circuits validateKey when the key matches the
allowlist. No per-user coordination needed, admin still controls
the policy (can rotate, can unset, can rate-limit).
## Changes
### server.mjs
* Line 100-103: new `PROXY_ANONYMOUS_KEY` env var + warning if set
while AUTH_MODE is not multi.
* Line 1127-1138 (localhost branch): anonymous allowlist short-circuit
before validateKey so localhost clients using the anon key are
labeled "anonymous" instead of "local" in stats.
* Line 1153-1163 (multi branch): the same allowlist check between
the ADMIN_KEY check and validateKey. Uses timingSafeEqual with
explicit length check (consistent with the admin and shared key
patterns above).
* Line 1221 (/health response): new `anonymousKey` field, null when
not set, the actual value when set. Admins opted into public
access when they set the env var, so exposing the key here is
intentional and lets ocp-connect auto-discover it without
out-of-band coordination.
### package.json
Version 3.6.0 to 3.7.0 (per dev iron rule 5, version bump before push).
### README.md
New "Anonymous Access (optional)" subsection under Auth Modes:
* Enable example (export PROXY_ANONYMOUS_KEY=...)
* Client-side /health discovery explanation
* Security note: opt-in to public access, rate-limit warning
* "Not a secret" note: /health is unauthenticated, the key is
publicly readable by design; treat it as a convenience handle,
not as an access credential.
## Test evidence
Offline tests on macOS 13, local server on 0.0.0.0:8889 with
PROXY_ANONYMOUS_KEY=ocp_public_anon_TEST, CLAUDE_AUTH_MODE=multi,
CLAUDE=/bin/echo (mock):
* GET /health ->
authMode: multi
anonymousKey: ocp_public_anon_TEST
(pass, field exposed correctly)
* POST /v1/chat/completions with Bearer ocp_WRONG_KEY from
172.16.2.29 (non-localhost) -> HTTP 401 in 15ms
body: Unauthorized: invalid or revoked API key
(pass, validateKey still rejects unknown keys)
* POST /v1/chat/completions with Bearer ocp_public_anon_TEST from
172.16.2.29 -> curl hangs at 3s after auth middleware
(pass, auth passed, request reached Claude handler which hangs
because CLAUDE=/bin/echo cant serve a real chat; the point is
the auth middleware accepted the key, confirmed by no 401 return)
* POST /v1/chat/completions with no Authorization from 172.16.2.29
-> curl hangs at 3s after auth middleware
(pass, pre-existing empty-token anonymous path not regressed)
node --check server.mjs: syntax OK.
## Code review
One independent opus reviewer. Verdict: PASS WITH CONCERNS, zero
blockers. Two strong-recommend items addressed in this commit:
1. Localhost branch was not covered in the first draft. Reviewer
pointed out that a localhost client using the anon key would be
labeled "local" instead of "anonymous" in stats, making operations
visibility worse. Fixed by applying the same anonymous allowlist
check in the localhost branch at line 1127-1138.
2. README security note was missing the explicit "not a secret"
framing. Added a paragraph clarifying that because /health is
unauthenticated, the anonymous key is publicly readable by
anyone who can reach the server, which is intentional.
Reviewer nits (tokenBuf3 naming, stats subcategory for header-less
vs anon-key anonymous) are deferred as they do not affect
correctness.
## Upstream dependencies on this commit
After this PR is merged and the OCP instance at 172.16.2.30 is
restarted with `PROXY_ANONYMOUS_KEY` set in the environment, a
follow-up PR can add anonymous key auto-discovery to ocp-connect:
* On startup, ocp-connect calls GET /health
* If anonymousKey field is non-null and --key was not provided,
ocp-connect automatically uses that key to seed per-agent
auth-profiles for OpenClaw
* User gets zero-config OCP connectivity for OpenClaw multi-agent
setups, no admin coordination, no --key flag
That follow-up is NOT in this PR. This PR is server-only.
Co-authored-by: taodeng <taodeng@Taos-MBP>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When the /api/oauth/usage endpoint returns 429 (rate limit), fall back
to the most recent cached data instead of returning an error. Also
extended cache TTL from 5min to 15min to reduce API calls.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds CLAUDE_NO_CONTEXT=true env var that passes
CLAUDE_CODE_DISABLE_CLAUDE_MDS=1 and CLAUDE_CODE_DISABLE_AUTO_MEMORY=1
to Claude CLI child processes. This suppresses CLAUDE.md and auto-memory
injection while preserving OAuth auth — needed for third-party agents
like Hermes that have their own memory systems.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replaces the old approach (sending a real messages API request just to
read rate-limit headers) with the dedicated usage endpoint that Claude
Code CLI uses. Fixes intermittent "unknown" plan usage.
- GET /api/oauth/usage with Bearer auth + anthropic-beta header
- Auto-refresh expired OAuth tokens via refresh_token grant
- Try both keychain label formats (claude-code-credentials, Claude Code-credentials)
- Zero API quota consumption for usage checks
- Parse new JSON response format (five_hour/seven_day structure)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Multi mode now allows unauthenticated requests as "anonymous".
Keys are optional: provide one for per-key usage tracking, or
skip it for zero-config access. Invalid keys are still rejected.
Localhost requests are always admin-level (can manage keys,
view dashboard data, etc.) without any token.
This makes OCP much friendlier for home LAN setups — family
members can use it immediately without any key configuration.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Requests from 127.0.0.1/::1 skip authentication even in multi/shared
auth modes. Only remote (LAN) clients need API keys.
This simplifies local tool integration — IDEs and agents on the same
machine no longer need to configure Bearer tokens.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Support ?token=xxx query parameter for dashboard auth (enables
headless browser screenshots and direct links)
- Token is saved to localStorage and URL is cleaned up
- Fix pathname matching to handle query parameters in URL
- Replace html2canvas screenshot with Playwright (accurate colors)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Fix 1 (keys.mjs): listKeys() no longer leaks full API key field in response
- Fix 2 (server.mjs): Remove BIND_ADDRESS admin bypass from isAdmin check
- Fix 3 (server.mjs): Admin key comparison now uses timingSafeEqual
- Fix 4 (server.mjs): Cap limit (max 500) and hours (max 720) in GET /api/usage
- Fix 5 (server.mjs): Streaming error path now computes promptChars from messages
- Fix 6 (server.mjs): Warn at startup if AUTH_MODE=shared but PROXY_API_KEY is empty
- Fix 7 (server.mjs): All recordUsage calls wrapped in try/catch to prevent DB errors crashing server
- Fix 8 (server.mjs): CORS fallback changed from "*" to specific origin (http://127.0.0.1:<PORT>)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Remove firstByteTimeout, adaptive tier algorithm, and MODEL_TIMEOUT_TIERS.
Claude tool-use causes 30s-5min pauses in token streams, making fine-grained
timeouts unreliable — they repeatedly killed valid requests. A single generous
timeout (10 min, matching LiteLLM/OpenAI SDK defaults) is simpler and correct.
Breaking: CLAUDE_FIRST_BYTE_TIMEOUT env var removed, default timeout changed
from 300s to 600s. Bump to v3.3.0.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The adaptive timeout calculation ignored BASE_FIRST_BYTE_TIMEOUT (set via
CLAUDE_FIRST_BYTE_TIMEOUT env), always using the tier base (120s for Sonnet).
Now uses whichever is larger: adaptive or env override. Bump to v3.2.2.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
On Linux (no macOS keychain), read ~/.claude/.credentials.json first.
Falls through to macOS keychain if file not found. Fixes /ocp usage
showing "No OAuth token" on Linux servers and RPi.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Reverts b5be1c0 and 1f1b387. These changes broke the OAuth token
authentication for fetching plan usage data from Anthropic API,
causing /ocp usage to show 0% with "unknown" reset times.
Root cause: code changes inadvertently altered the fetchUsageFromApi
flow, breaking the existing working OAuth x-api-key authentication.
This restores server.mjs to the last known working state (f6f4f68).
Phase 1 modular files removed — will be re-implemented in a separate
branch with proper isolation testing.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When the Claude Code OAuth token expires, /ocp usage showed 0% for all
plan limits. OCP now detects expiry and triggers a refresh via Claude CLI,
then re-reads the updated token from keychain. The 5-minute usage cache
prevents excessive refresh attempts.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Emergency fix for multi-agent (ClawTeam) burst scenario that caused cascading
failures: when multiple Opus agents ran concurrently, the old consecutive-count
breaker (threshold=3) tripped within seconds, blocking ALL subsequent requests
globally — including non-ClawTeam agents and new sessions.
Changes:
- Circuit breaker now uses a sliding time window (default 5min) instead of
consecutive failure count. 6 failures in 5min triggers open, vs old 3-in-a-row.
- Half-open state allows 2 concurrent probe requests (was 1), improving recovery
speed when API comes back.
- Graduated backoff: cooldown doubles on each re-open (120s→240s→300s cap),
resets fully on first success.
- Increased default timeouts: Opus first-byte 60s→90s, Sonnet 45s→60s,
overall 120s→300s, max concurrent 5→8.
- /health endpoint now exposes per-model breaker state, window stats, and
marks status as "degraded" when any breaker is open.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* Add graceful shutdown handling for SIGTERM and SIGINT
On receiving SIGTERM or SIGINT, the server now:
1. Stops accepting new connections (server.close)
2. Clears session cleanup and auth check intervals
3. Sends SIGTERM to all active child processes
4. Waits up to 5s for processes to exit, then SIGKILL + exit(1)
5. Exits cleanly with exit(0) once all processes are gone
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: security hardening — bind localhost, validate models, limit body size
- Bind to 127.0.0.1 instead of 0.0.0.0 to prevent network exposure
- Restrict CORS Access-Control-Allow-Origin to http://127.0.0.1
- Add 10MB request body size limit (413 on exceed)
- Validate model parameter against known MODEL_MAP keys (400 on unknown)
- Remove catch-all POST route; only /v1/chat/completions accepted
- Use crypto.timingSafeEqual for API key comparison
- Sanitize error messages to strip internal file paths
Bumps version to 2.4.1.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: replace fake streaming with real stdout-piped SSE streaming
Previously, streaming requests waited for the entire claude CLI process
to complete, then split the output into artificial 500-char chunks sent
as SSE events. This defeated the purpose of streaming and added latency.
Now, stdout from the claude child process is piped directly to SSE
chunks as data arrives. Each `data` event from the process stdout
becomes an immediate SSE delta event, giving clients real incremental
output. The shared process setup logic is extracted into
spawnClaudeProcess() to avoid duplication between streaming and
non-streaming paths. Client disconnect detection kills the child process
to free resources.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: tighten CORS to dynamic localhost-only and reduce body limit to 5MB
- CORS now dynamically matches localhost/127.0.0.1 origins instead of
static http://127.0.0.1, preventing cross-origin abuse while
supporting any localhost port
- Body size limit reduced from 10MB to 5MB to limit OOM risk
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Oracle Public Cloud User <opc@instance-20230820-1333.subnet07301351.vcn07301351.oraclevcn.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
- Circuit breaker: consecutive timeouts (default 3) mark model as degraded for 60s,
failing fast instead of blocking gateway with repeated timeout attempts
- Per-model timeout tiers: Opus 60s base, Sonnet 45s, Haiku 30s, plus adaptive
scaling by prompt size (~15s/100k chars for Opus)
- Structured JSON logging for spawns, timeouts, breaker state changes
- Late close guard: prevents double-settle when timeout fires before proc.close
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Reduced default CLAUDE_TIMEOUT from 300s to 120s for faster fallback
- Added CLAUDE_FIRST_BYTE_TIMEOUT (default 30s): aborts early if Claude
CLI produces no output, preventing silent hangs
- First-byte timing logged for every request for observability
- Health endpoint now reports firstByteTimeout in config
Major architectural upgrade:
- Replace pool system with on-demand spawning (eliminates crash loops,
DEGRADED states, and stdin timeout errors from v1.x)
- Add session management with --resume support (reduces token waste on
multi-turn conversations)
- Expand default allowed tools (Bash, Read, Write, Edit, Glob, Grep,
WebSearch, WebFetch, Agent) with configurable CLAUDE_ALLOWED_TOOLS
- Add system prompt pass-through (CLAUDE_SYSTEM_PROMPT)
- Add MCP config support (CLAUDE_MCP_CONFIG)
- Add concurrency control (CLAUDE_MAX_CONCURRENT, default 5)
- Add periodic auth health monitoring
- Add session API endpoints (GET/DELETE /sessions)
- Improve /health with full diagnostics (stats, sessions, errors, config)
- Fix timeout race condition (graceful SIGTERM → SIGKILL)
- Fix ERR_HTTP_HEADERS_SENT by checking headersSent in all response helpers
- Document coexistence with Claude Code interactive mode (Telegram, IDE)
- No conflict with CC: different ports, protocols, and process models
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The proxy previously defaulted to bare "claude" command which fails in
macOS LaunchAgent where PATH is minimal. Now resolves the binary at
startup via: CLAUDE_BIN env → well-known paths → which fallback.
Fails fast with clear error if binary cannot be found.
Also enhances /health to report claudeBinary and claudeBinaryOk status.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add optional PROXY_API_KEY env var for Bearer token auth
- Return 401 for missing/invalid token on all endpoints except /health
- Skip auth when PROXY_API_KEY is not set (backwards compatible)
- Log auth status on startup
- Rework replenishPool to accept isCrash flag: initial fills spawn
immediately with no delay; only crash-triggered respawns apply backoff
- Exponential backoff on crash: 2s, 4s, 8s, 16s, 32s, capped at 60s
- After 5 consecutive fast crashes (< 10s lived) within a 60s rolling
window, mark model as "degraded" and stop all respawning to prevent
CPU spin loops
- Degraded state resets automatically if a subsequent spawn lives > 10s
- stderr from crashed pool workers is captured and logged on exit
- Bump version to 1.6.0
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
