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
Commits on main
Merge pull request #120 from gglucass/fix/mcp-tool-name-prefix
47e734c Tejas Chopra committed 7 months ago
Support both API key and codex login auth for Codex WS proxy
059e165 chopratejas committed 7 months ago
Add INFO-level logging for WS /v1/responses connect and complete
92f74db chopratejas committed 7 months ago
Auto-inject Codex provider config with supports_websockets=true
bd948b4 chopratejas committed 7 months ago
Count Strands reasoningContent, image, document, video tokens (#111 follow-up)
e7008f6 chopratejas committed 7 months ago
Fix /v1/responses/* passthrough for ChatGPT session auth (#71)
c92a6b9 chopratejas committed 7 months ago
Add passthrough route for /v1/responses/* sub-endpoints (compact, cancel, etc.)
532d9da chopratejas committed 7 months ago
Add sentencepiece to image extra — required by SigLIP tokenizer (#103)
c52b0f1 chopratejas committed 7 months ago
Fix Strands SDK tool pairing, CCR marker format, and startup timeout (#114 #116 #117)
d7122d6 chopratejas committed 7 months ago
Merge pull request #112 from JerrettDavis/feat/docker-e2e-wrap-validation
3e94e53 Tejas Chopra committed 7 months ago
Merge upstream/main into feat/docker-e2e-wrap-validation
addcfcf JerrettDavis committed 7 months ago
chore: fix ruff lint issues (import sorting, trailing whitespace)
4baef0e Tejas Chopra committed 7 months ago
fix: eliminate duplicate proxy logging and guard against token inflation
039399a Tejas Chopra committed 7 months ago