11 Commits

Author SHA1 Message Date
amos
b773033658 Record model requests and payloads as prefix deltas in the native trace
Each model_request/model_payload repeated the full context, so the trace
grew with the square of the request count: batch 6 Qwen runs hit 64 MiB
at ~75 requests and 14 attempts (13 graded PASS) were discarded.
Requests now record only items after the common prefix with the previous
one (schema embedded-rollout.pi/2), and pi_run.expand rebuilds full
payloads before the unchanged broker/feedback checks. Needs a new image
(trace6) and recalibration before use.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-18 10:50:56 +08:00
amos
b71d605d81 Lower the shared-disk floor to 30 GiB at the user's request
The other users of /home left 57 GiB free; the user chose to keep running
and stop only at 30 GiB. Batch start and runtime disk_low now share that floor.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-17 13:11:21 +08:00
amos
141f3d06aa fix: cap disk use per attempt and refuse to start on a low output disk
Agent containers wrote to bind-mounted /workspace and /trace with no
size limit, so a runaway Agent could fill the shared /home. Containers
no longer start below 50 GiB free; a running Agent is stopped when the
disk drops below that floor or its attempt directory passes 2 GiB.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-14 15:22:35 +08:00
amos
c6435255f9 docs: record the user-specified Git remote
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-13 20:49:41 +08:00
amos
f0c277a0d1 fix: controller owns model output limits; verify them on the wire
Output limits were baked into pi.mjs inside the runtime image, so every
change (GLM 32000 -> 131072, Qwen 8192 -> 65536) needed a new image and a
full 785-trial recalibration, and the Qwen baseline ran on an image the
repo could not rebuild.
- MODELS holds max_tokens and its wire field per model; pi_run passes
  --model and --max-tokens to the adapter, which has no defaults left.
- Receipts record model_max_tokens and model_sampling=service-default;
  evidence requires every payload to carry exactly that limit, which
  would have caught the SDK's silent 32000 cap.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-13 20:27:29 +08:00
amos
7071ecfa7f fix: raise the native trace limit to 64 MiB
Bos tasks with 131072-token reasoning produce ~110 bytes per streamed
token plus a full context per request; the trace3 batch hit 16 MiB
before finishing. Adapter and host share the new limit, and the size
tests now derive their data from MAX_TRACE_BYTES. Runtime trace4 carries
this adapter; all 21 tasks recalibrated.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-12 23:05:39 +08:00
amos
1b73b3742b fix: keep host stdout complete and lift the GLM output cap
The first Bos pilot batch exposed two adapter faults on long responses:
- fs.writeSync(1, ...) ignored partial writes to a nonblocking stdout pipe,
  so host stdout lost the tail of 100 KB lines and could no longer match
  the native trace. writeAll now loops until every byte is written and
  waits briefly on EAGAIN.
- Pi SDK caps max_tokens at 32000 unless told otherwise; GLM-5.3 spent
  all of it reasoning and stopped with "length" before any edit. The
  adapter now passes the registered 131072 (accepted upstream).

Runtime trace3 is the trace2 digest plus this pi.mjs; 12/12 SDK checks
pass inside it and all 21 tasks were recalibrated.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-12 22:10:12 +08:00
amos
acfa601a56 feat: multi-file task contract, model table and shared I/O helpers
- Tasks may declare up to 16 candidate files; the whole template becomes
  the Agent workspace while only candidate files are frozen and graded.
- Optional fixtures add more cheats or legal alternatives (five trials
  each); optional grader_files ship private tests next to verify.py.
- Tasks without an argv oracle export call_type=verifier ground truth.
- Upstream model routes live in one MODELS table.
- One bounded no-follow reader, one exclusive writer and one SIGTERM
  helper replace per-module copies; drop unused validate_task.

frame-decoder keeps its task digest, calibration receipts and exported
query. Verified on host 36: internal network probe, 20-trial calibration
and a real GLM rollout (PASS, evidence VERIFIED, cleanup confirmed).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-12 21:25:09 +08:00
amos
e524d9eda5 docs: deployment map, task authoring standard and Qwen trial record
Add the current and target deployment diagrams, the task author standard
and copyable spec template, the five-field data contract, and record the
Qwen Responses trial and pending-create cleanup fix in progress and
acceptance.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-12 21:06:06 +08:00
amos
d205c44249 fix: cut agent internet egress and loosen grader timeouts
The dedicated Pi network was a plain bridge, so the Agent could reach the
internet and fetch upstream fixes. Create it with --internal, make run
accept only internal networks, and default to --network none.

Verifier limits of 1s per case and 15s per grade turned host load into
false FAIL verdicts. Use 5s per case, 30s compile, 10s isolation probe and
timeout_seconds=60; frame-decoder task version becomes 2.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-12 20:30:00 +08:00
amos
65ceed6c46 docs: establish rollout scope and development contracts 2026-10-12 11:15:44 +08:00