The first judge-spend rows recorded by 248948c came back with input_tokens = 0
on both passes of mission 01a09dfc — 7 and 9 requests, 5940 and 2109 output
tokens, and nothing on the side that actually empties the plan. Probed z.ai's
Anthropic-compatible stream directly: `message_start` carries
`"input_tokens": 0`, and the real figure arrives in `message_delta.usage`
beside output_tokens. Anthropic proper does it the other way round, which is
the shape the parser was written for.
A nonzero figure in the delta now wins; otherwise the start's figure stands,
so the Anthropic path is byte-for-byte unchanged. The decision is a pure
function with the three shapes as its test — including a delta that says 0,
which must not erase what the start said.
Co-Authored-By: Claude Opus 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01WZb5A2kfVfjpdwSochkuHz
On the OAuth path Anthropic requires the Claude Code identity to be its own
first system BLOCK. We concatenated it with the caller's prompt into a single
string, so EVERY server-side call that set a system prompt was rejected — and
the rejection arrives as `429 {"type":"rate_limit_error","message":"Error"}`,
which reads as throttling and is not.
Measured on one token, seconds apart:
"PREAMBLE" (string) -> 200
"PREAMBLE\n\nJudge the …" (string) -> 429
"PREAMBLE" (string) -> 200 (control)
["PREAMBLE"] (blocks) -> 200
["PREAMBLE", "Judge the …"] (blocks) -> 200
while the account reported `5h utilization 0.07, 7d 0.11, overage 0.0`, every
window `allowed`. A Max 20x subscription at 7% was being read as out of
capacity.
What this was breaking, silently, for as long as it has been there:
- every `done_when` verdict on the subscription judge. Mission 01a00bbb
pass 2 returned "could not evaluate the completion condition this pass"
and BURNED one of the phase's three passes on it.
- the boot preflight, which reported `claude-opus-4-8 throttled (configured,
no capacity now)` on every start — a diagnostic that was itself the bug.
- mission_refiner, phase_summarizer, swarm planning.
The `claude` CLI was unaffected throughout, because it sends its system prompt
as blocks. That divergence is what made this look like an account problem: the
agents worked while everything server-side "throttled".
After the fix the preflight reports opus-5, sonnet-5 and haiku all `ok`.
The API-key path keeps sending a plain string — it never had this constraint.
Co-Authored-By: Claude Opus 5 <[email protected]>
A phase verdict is a classification: fixed prompt, no tools, no memory, one
JSON answer. Routing it through a ZeroClaw agent charged 17,772 input tokens
to produce a 20-token reply, and at the runtime's 32k context that scaffolding
— role prompt, tool descriptors, memory, identity — consumed over half the
window before the judge read any evidence.
The same verdict as a direct Messages API call costs 114 input tokens, with
the real system prompt and evidence. Measured through the production seam via
`cargo run -p cm-llm --example oauth_probe`.
- cm-llm: teach AnthropicProvider subscription auth. A `sk-ant-oat…`
credential switches to bearer auth, adds the Claude Code beta set, and
prepends the identity line the API requires as the first system block —
idempotently, so re-wrapping can't stack it or waste tokens.
- evaluator: prefer a direct provider call whenever ANTHROPIC_OAUTH_TOKEN is
set, falling back to the configured spec (including `runtime:<alias>`)
otherwise. Fail-closed parsing is untouched and still governs every path.
- The ANTHROPIC_API_KEY shape guard now points at the slot that understands
bearer auth rather than only saying no.
Deleting the agent from this path is the ablation applied to our own harness:
the scaffolding was there because a judge was built like every other agent,
not because a judge needs it.
Co-Authored-By: Claude Opus 5 <[email protected]>
Running the opt-in live suite (CM_LIVE_LLM=1 + ANTHROPIC_API_KEY) against
the real API immediately surfaced a launch blocker: Anthropic (and
OpenAI) restrict tool names to ^[a-zA-Z0-9_-]{1,128}$ — our ENTIRE
registry uses dotted names (clock.now, email.send, shell.exec, ...).
The scripted provider never enforced the pattern, so every real-model
deployment would have 400'd on the first tool call.
- Fix at the provider boundary, where it belongs: wire_tool_name /
internal_tool_name codec (dots <-> __) applied in BOTH HTTP providers
at all three sites (tools list, assistant tool_use echo, inbound
tool_use decode). Internal naming (DB step rows, scenarios, UI traces)
unchanged. Offline unit test round-trips every registry name through
the wire pattern
- New live tests, all passing against api.anthropic.com (Haiku 4.5):
- provider tool ROUND TRIP: real ToolUse arrives, ToolResult ships
back exactly as a checkpoint would reassemble it, model completes,
real usage events on the wire
- full runtime loop: real model calls clock.now, run completes, REAL
token usage metered, credits decremented
- the #1-risk validation: a real model's email.send intercepted ->
suspended -> approved -> checkpoint RESUMED against the live API ->
completed -> outbox exactly 1 (checkpoint/resume fidelity end to end)
- Stray TC_OPENAI_COMPAT_* envs renamed to CM_OPENAI_COMPAT_*
No credentials stored anywhere; the key was passed via env only.
163 Rust tests (+5 live, key-gated).
Co-Authored-By: Claude Fable 5 <[email protected]>
Full-depth rename per the approved plan; the 'claw' product vocabulary
(claws, /claws routes, clawId, Claw Chat) stays — it is now the brand.
- Display brand: Clawmates (manifest, titles, hero, login/rail logo
'clawmates'); default host app.clawmates.work; registry
ghcr.io/clawmates
- Crates tc-* -> cm-* (16 crates + all imports); binaries
clawmates-server/broker/bundler; images clawmates/*; env prefix
CLAWMATES_* (+ CM_TEST_DATABASE_URL / CM_LIVE_LLM); config
clawmates.toml; helm chart deploy/helm/clawmates with clawmates-*
resources; db names clawmates*; sockets /run/clawmates; cookie
cm_session; kind cluster clawmates-test; seccomp node profile
clawmates-agent-profile.json
- All 9 Playwright brand assertions updated in lockstep; historical
spec document left untouched as the only remaining 'TeamClaw'
- Local env migrated: dev pg clawmates-dev-pg/clawmates_dev, shared
test server clawmates-test-pg, kind cluster recreated with image +
profile, compose images rebuilt under clawmates/*
Verified end to end: 161 Rust + 68 frontend tests, 29 Playwright
journeys, 4 live kind tests, helm/install/LOC/placeholder gates, and
the clean-room install rehearsal serving the clawmates login page from
a signed bundle of the rebuilt images.
Co-Authored-By: Claude Fable 5 <[email protected]>
