Skip to content

Commit 957eeee

Browse files
sjarmakclaude
andcommitted
Fix openlibrary-solr-boolean-fix-001 baseline: pre-install Node.js + Claude Code
The SWE-bench Pro base image for this task has broken NodeSource GPG (needs /dev/tty which doesn't exist in containers), preventing Harbor's install.sh from installing Node.js → Claude Code. Fix: pre-install Node.js 22 via tarball + Claude Code via npm at Docker build time. If claude is already on PATH, Harbor's claude_run.sh finds it and skips the broken NodeSource install. Verified: agent now runs successfully (174-line transcript vs previous 1-line "claude: command not found"). Baseline reward=0.0 (legitimate agent failure, not infra error). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
1 parent 9834b8d commit 957eeee

File tree

3 files changed

+19
-1
lines changed

3 files changed

+19
-1
lines changed

benchmarks/ccb_fix/openlibrary-solr-boolean-fix-001/environment/Dockerfile

Lines changed: 9 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -4,6 +4,15 @@
44

55
FROM jefzda/sweap-images:internetarchive.openlibrary-internetarchive__openlibrary-92db3454aeaa02f89b4cdbc3103f7e95c9759f92-v2c55207218fb8a0138425cbf7d967
66

7+
# Pre-install Node.js 22 + Claude Code at build time.
8+
# Harbor's install.sh fails on this image (NodeSource GPG needs /dev/tty)
9+
# but if claude is already on PATH, /tmp/claude_run.sh finds it.
10+
RUN curl -fsSL https://nodejs.org/dist/v22.14.0/node-v22.14.0-linux-x64.tar.gz \
11+
| tar -xz -C /usr/local --strip-components=1 \
12+
&& node --version && npm --version \
13+
&& npm install -g @anthropic-ai/claude-code@latest \
14+
&& which claude && claude --version
15+
716
# Install uv for Python package management
817
RUN curl -LsSf https://astral.sh/uv/0.7.13/install.sh | sh || true
918

docs/ops/SCRIPT_INDEX.md

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -190,6 +190,7 @@ Generated from `scripts/registry.json` by `scripts/generate_script_index.py`.
190190
- `scripts/rerun_errored_tasks.sh` [one_off] - Historical one-off script: rerun errored tasks.
191191
- `scripts/rerun_fixed_tasks.sh` [one_off] - Historical one-off script: rerun fixed tasks.
192192
- `scripts/rerun_zero_mcp_tasks.sh` [one_off] - Historical one-off script: rerun zero mcp tasks.
193+
- `scripts/rescore_difficulty.py` - Utility script for rescore difficulty.
193194
- `scripts/run_judge.py` - Utility script for run judge.
194195
- `scripts/scan_swebench_errors.py` - Utility script for scan swebench errors.
195196
- `scripts/smoke_artifact_verifier.py` - Utility script for smoke artifact verifier.

scripts/registry.json

Lines changed: 9 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -962,6 +962,14 @@
962962
"language": "shell",
963963
"summary": "Historical one-off script: rerun zero mcp tasks."
964964
},
965+
{
966+
"name": "rescore_difficulty.py",
967+
"path": "scripts/rescore_difficulty.py",
968+
"category": "misc",
969+
"status": "maintained",
970+
"language": "python",
971+
"summary": "Utility script for rescore difficulty."
972+
},
965973
{
966974
"name": "retrieval_eval_pipeline.py",
967975
"path": "scripts/retrieval_eval_pipeline.py",
@@ -1171,7 +1179,7 @@
11711179
"infra_mirrors": 16,
11721180
"library_helpers": 7,
11731181
"migration": 4,
1174-
"misc": 39,
1182+
"misc": 40,
11751183
"qa_quality": 10,
11761184
"submission_reporting": 7,
11771185
"task_creation_selection": 12,

0 commit comments

Comments
 (0)