Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
Python agentaianthropicclaude-codecompressioncontext-engineeringcontext-windowcursorfastapilangchainllmmcpopenaiprompt-engineeringproxypythonragtoken-optimizationtokenstypescript
fix(ci): guarantee model present in test shards to end cache-miss flakiness (#1399)
JD Davis2e29c7223f7a
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.ymlindex 7168250af..b2a542f25 100644--- a/.github/workflows/ci.yml+++ b/.github/workflows/ci.yml@@ -58,6 +58,7 @@ jobs:- 'Cargo.lock'- 'tests/**'- 'scripts/**'+ - '.github/workflows/**'e2e:- 'headroom/**'- 'crates/**'@@ -170,11 +171,27 @@ jobs:restore-keys: ${{ runner.os }}-pip-${{ env.PY_VERSION }}-- name: Restore HuggingFace model cache (warmed by prefetch-model)+ id: restore-hfcacheuses: actions/cache@v5with:path: ~/.cache/huggingfacekey: ${{ runner.os }}-models-allMiniLM-v2+ - name: Fallback model download if cache missed+ if: steps.restore-hfcache.outputs.cache-hit != 'true'+ env:+ HF_TOKEN: ${{ secrets.HF_TOKEN }}+ HF_HUB_DISABLE_TELEMETRY: "1"+ TRANSFORMERS_OFFLINE: "0"+ HF_HUB_OFFLINE: "0"+ run: |+ python -m pip install --upgrade pip huggingface_hub+ for i in 1 2 3 4 5 6; do+ if python -c "from huggingface_hub import snapshot_download; snapshot_download('sentence-transformers/all-MiniLM-L6-v2')"; then exit 0; fi+ if [ "$i" -lt 6 ]; then echo "::warning::fallback model fetch attempt $i failed; backing off"; sleep $((i * 30)); fi+ done+ echo "::error::could not fetch all-MiniLM-L6-v2 from HuggingFace (fallback)"; exit 1+- name: Download prebuilt wheeluses: actions/download-artifact@v8with: